EPA Contract No. 68-03-3268
Work Assignment No. 1-10
FEASIBILITY OF CHROMIUM SPECIATION FOR EPIDEMIOLOGICAL EXPOSURE MODEL
Submitted to:
U.S. Environmental Protection Agency
Office of Research and Development/ECAO
Cincinnati, Ohio 45268
Project Officer: Linda Schwaegerle
Technical Project Monitor: Herman Gibb
Submitted by:
Technical Resources, Inc.
3202 Monroe Street, Suite 300
Rockville, Maryland 20852
Prepared by:
Peter S.J. Lees, Ph.D., C.I.H.
The John Hopkins University
School of Hygiene and Public Health
Department of Environmental Health Sciences
Division of Environmental Health Engineering
615 North Wolfe Street
Baltimore, Maryland 21205
-------
EXECUTIVE SUMMARY
A summary of the existing database to be utilized in the modeling of
chromium (Cr) exposures at the Baltimore Chromate Works of the Allied
Corporation is presented. Description and discussion focus on documentation
of exposures to various Cr species in response to EPA's request to evaluate
the feasibility of isolating risks by disease by Cr species. Data from the
Allied Corporation, U.S. Public Health Service, Baltimore City Health
Department and Dr. Anna Baetjer's archives are reviewed. Although all records
have not been collected or examined, it appears that the U.S. Public Health
Service study (ca. 1952) contains the only direct measures of employee
exposure which differentiate exposure by Cr species. More recent exposure
data differentiating chromium species are not available; all samples have
apparently been analyzed for Cr content only.
Although the existing database contains limited information about worker
exposures by Cr species, it may be possible to partition historical exposures
by species. It is recommended that the existing epidemiological data be
re-analyzed to assess the difference in disease incidence (if any) between
workers in the mill and roast plant (with a mixed Cr /Cr exposure) and
the special products/acid plant (with a pure Cr exposure) to assess the
separate risks of disease. In addition it is recommended that historical
process quality control data and settled dust samples be utilized to estimate
quantitative levels of exposure Cr for inclusion in the Cr exposure
model.
-------
TABLE OF CONTENTS
I. PURPOSE/OBJECTIVES 1
II. DESCRIPTION AND DISCUSSION OF EXISTING DATA
A. Allied Corporation Data 1
B. U.S. Public Health Service Data 3
C. Baltimore City Health Department Data 4
D. Dr. Anna Baetjer's Archives 4
III. FEASIBILITY ASSESSMENT/CONCLUSIONS 4
IV. RECOMMENDATIONS 6
-------
I. PURPOSE/OBJECTIVES
This report has been prepared to describe, in general, the existing
database to be utilized in the modeling of chromium exposures at the Baltimore
Chromate Works of the Allied Corporation. The following description and
discussion focus on documentation of exposures to various chromium (Cr)
species in response to EPA's specific request to determine the feasibility of
utilizing the database to isolate the risk of disease (if any) from exposure
to Cr and the generally recognized risk of disease from exposure to
Cr . As the environmental exposure database has not been fully assembled
at this writing, the discussions and conclusions may be somewhat premature and
are, therefore, subject to change as additional data are gathered.
II. DESCRIPTION AND DISCUSSION OF EXISTING DATA
A. Allied Corporation Data
To date several meetings have been held with representatives of the
Allied Corporation to define the data required for this study. The Allied
Corporation has been very receptive to our requests and continues to be very
co-operative. Because of a major change in administrative structure and
personnel during the course of these discussions, however, the actual transfer
of the Allied exposure data to the project investigators has been delayed.
The following description of the existing database is based on discussions and
meetings with the last chief industrial hygienist at the plant before it
closed, and with the current director of environmental operations at the plant
who is responsible for, among many other things, the collection and archiving
of all records related to the plant.
The plant which is the subject of this study was constructed in 1952. It
is a large industrial complex, but can be divided into two basic parts: the
mill and roast department and the special products/chromic acid departments.
Only monor changes were made in the mill and roast portions of the plant
between 1950 and the plant closing in 1985. The special products and chromic
acid portion of the plant underwent major renovations in 1960-61. Both parts
of the plant are located in separate buildings which are large open structures
in which personnel moved about freely during the course of a day. The
-------
physical separation of these two plants and the frequent crossover of the
working populations has important implications for measurement of the risk of
disease by chromium species. Exposure in the mill and roast department was to
O I s i
a mixed Cr /Cr dust while in the special products/chromic acid
departments exposure was limited to a pure Cr dust.
From 1952 until 1985 the plant operated, with very few interruptions, at
essentially 100% of capacity. Process quality control data associated with
production records are a potential source of exposure speciation information.
Investigation to date has shown that process quality control measurements were
probably limited to soluble chromium compounds and iron at various points in
the manufacturing process. Insoluble chromium species were not normally
measured as a part of quality control activities. Insoluble species may,
however, be estimated from the data by difference at each stage of the
process. The location and state of the old process quality control records is
not known at this time; there is, however, the possibility that additional
information exists.
The Allied Corporation bought the Mutual Chemical Company Chromate Works
in the late 1950's (?) and inherited all records from the plant dating back to
at least the last century. The Allied Corporation and its predecessor have
collected environmental exposure data at least from the late 1940's. These
data were collected by the process engineering department until approximately
1970 when a separate department for health and environment was formed. The
environmental data for the 1940-70 era are not currently stored at the
Baltimore Works and, due to delays caused by administrative changes, have only
been positively located in the last week.
Environmental exposure data from the 1940-70 era are said to be
considerable (perhaps 100 samples per year on average) but have not been
examined to date. The quality of the data has, therefore, not be assessed.
It is the recollection of plant personnel that, although exposure
concentrations are available, there is little documentation as to sampling and
analytical methods employed, measurement units are often missing, etc.
In the post-1970 era extensive environmental exposure data were
collected. From 1970 until the plant closed in 1985, area air samples were
collected on a monthly basis from a standardized grid system within the plant.
-------
This grid system was modified in i960 to a system of twenty sampling zones.
Approximately 1500 samples were collected per year as part of these systems.
Beginning in approximately 1978, these area samples were supplemented with
personal samples of exposure. A total of some 1500 personal samples were
collected before the plant closed. The extensive and complete nature of the
data greatly facilitate the development of a general Cr exposure model for
workers at the plant.
All personal and area samples collected by Allied and Mutual at least
since the 1940's were only analyzed for Cr content; Cr and/or total Cr
were never measured on a regular basis and, to the extent permitted by our
limited examination of the data, were probably never collected. The sampling
and analytical methods employed in the 1940-70 era are not known at this time,
although it is generally believed that RAC tape samplers were employed
extensively in the 1960's. Although documents have not yet been found to
verify recollections, it is believed that all samples during this period were
analyzed for Cr only. The ai
known positively at this time.
analyzed for Cr only. The analytical method employed, however, is not
In the post-1970 period samples were collected by RAC tape samplers (area
samples) and personal sampling pumps and filters (peronal samples).
Apparently all of these samples were analyzed utilizing the diphenylcarbazide
method. This is a widely used standard method for the analysis of Cr
The Allied Corporation, however, modified the method somewhat by utilizing
acidified diphenylcarbazide instead of the sequential addition of
diphenylcarbizide and acid specified in the method. The bias introduced by
this variation (if any) will have to be independently evaluated. No direct
3+
information on Cr or total Cr exposure is, thus, available for this era.
At least some of the post-1980 exposure data are said to be available on
computer tape, but have not seen located as of this writing.
B. U.S. Public Health Service Data
Copies of the original raw data from the U.S. Public Health Service study
of the plant in the early 1950's have been obtained as has a copy of the
resulting analysis of the data published as "Health of Workers in ... Chromate
Producing Industry" (Public Health Service Publication No. 192). This study
-------
contains measures of air concentrations, settled dusts, and processes analyses
of various forms of chromium. In most instances, total Cr, Cr , soluble
and insoluble Cr concentration data are presented. The sampling and
analytical methods employed are well documented as a part of the final
report. The data contained in this study can serve as a solid basis for
estimation of Cr exposure concentrations by species for the early 1950's.
C. Baltimore City Health Department Data
The Bureau of Industrial Hygiene of the Baltimore City Health Department
made frequent measurements of employee exposures at the chromate plant from at
least the mid-1930's. Records of several hundred exposure measurements are
available in the Health Department's files. To date, these records have not
been scrutinized because a confidentiality agreement to protect individuals
named in the exposure records has not been reached between the Health
Department and the investigators.
D. Dr. Anna Baetjer's Archives
Dr. Anna Baetjer was closely associated with health- and exposure-related
studies at the chromate plant from the late 1940's until her death in 1984.
At this time it is known that her extensive records of this association
contain detailed observations of processes, controls, work practices, etc. at
the plant. These records will be extremely valuable in helping to interpret
the quanitative exposure data. It is believed at this time that Dr. Baetjer
did not conduct any quantitative exposure measurements herself, and that all
such data In her records may be found in its original form elsewhere. This
assumption, however, is unverified to date.
III. FEASIBILITY ASSESSMENT/CONCLUSIONS
Although all records have not been collected or examined, It appears that
the U.S. Public Health Service study contains the only direct measures of
employee exposure which differentiate exposure as to chromium species. These
data were all collected in the early 1950's. More recent exposure data
differentiating chromium species are not available; all samples have
apparently been analyzed for Cr content only.
-------
3+
Several methods of estimating Cr exposure and risk of disease are
possible utilizing the existing database. The most direct method of
differentiating disease risk by species would be to analyze existing (and
updated) vital status data treating employees of the mill and the roast plant
and the special products/acid plant as separate populations. As stated
earlier, exposure in the mill and roast department was to a mixed Cr /
Cr dust while in the special products/chromic acid departments exposure
was limited to a pure Cr dust. Independent analysis of the risk of
disease in these two plant populations, while probably requiring extensive
adjustments to make the populations otherwise comparable, would have the
direct advantage of showing a differences in risk which could be attributed to
the differences in exposure. Quantification of those exposures, however, is
not possible as a part of this type of study.
In the absence of direct measures of worker exposure by Cr species, the
most promising source of quantitative information may be derived from process
quality control data. These measurements, made on a regular basis at numerous
points in the manufacturing process, provide a direct measure of soluble Cr
content of the product. It may be possible to deduce insoluble Cr content by
calculating the difference in soluble Cr concentration from the preceding
step. The use of process quality control information to estimate worker
exposure by Cr species has two major implicit assumptions: 1) it must be
assumed that the partition of Cr species in the process materials was mimicked
in the dust in the air and 2) it must be assumed that the soluble and
insoluble fractions correspond in some fashion to the Cr and Cr
species. Specific Information may be present in the existing data set and/or
laboratory simulations could be designed to substantiate or refute these
assumptions.
A second possible source of quantitative information on the Cr species of
workers exposures would be the analysis of settled dust samples. This was
done as a part of the U.S. Public Health Service study In the early 1950's and
could be duplicated today. The use of these data to segregate worker
exposures by species would involve two assumptions similar to those required
above: 1) a straight line relationship would have to be assumed to explain
any changes in the partition of the Cr species between the Public Health
-------
Service measurements of the early 1950's and today (In addition, methods would
have to be carefully cross-calibrated) and 2) it must be assumed that the
partition of Cr species in the settled dust was mimicked in the dust inhaled
by the workers.
IV. RECOMMENDATIONS
Although the existing database contains only very limited (and old)
information about worker exposures by Cr species, it may nonetheless be
possible to partition historical exposures by species with a fair degree of
reliability. It is recommended Initially that the existing epidemiological
data be re-analyzed to assess the difference in disease incidence (if any)
between workers in the mill and roast plant (with a mixed Cr /Cr
exposure) and the special products/acid plant (with a pure Cr exposure) to
assess the separate risk of disease.
In addition it is recommended that if the epidemiological study outlined
above shows any hint of risk associated with Cr exposure that analysis of
historical process quality control data and settled dust samples be undertaken
to quantitate the level of exposure to Cr for inclusion in the Cr
exposure model. Due to time considerations, however, it may be desirable to
3+
begin efforts to quantitate Cr exposures immediately.
------- |