TEXTILE DYEING PLANT DRUG ROOM
MONITORING STUDY
Prepared by
Dr. Arnold Greenland and Dr. David Cox
THE WASHINGTON CONSULTING GROUP, INC.
1625 I Street, N.W., Suite 214
Washington, D.C. 20006
Mr. Don Harbin
MIDWEST RESEARCH INSTITUTE
425 Volker Boulevard
Kansas City, Missouri 64101
and
Ms. Paula A. Morelli-Schroth
PEI ASSOCIATES, INC.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68-02-4229
August 1988
Volume 2, APPENDICES
EPA Task Manager: Margaret Conomos
EPA Project Officer: Philip E Robinson
Prepared for:
Design and Development Branch
Exposure Evaluation Division (TS-798)
Office of Toxic Substances
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
WCG
The Washington Consulting Group
1625 Eye Street, N.W.
Suite 214
Washington, D.C. 20006
(202) 457-0233
-------
TEXTILE DYEING PLANT DRUG ROOM
MONITORING STUDY
Prepared by
Dr. Arnold Greenland and Dr. David Cox
THE WASHINGTON CONSULTING GROUP, INC.
1625 I Street, N.W., Suite 214
Washington, D.C. 20006
Mr. Don Harbin
MIDWEST RESEARCH INSTITUTE
425 Volker Boulevard
Kansas City, Missouri 64101
and
Ms. Paula A. Morelli-Schroth
PEI ASSOCIATES, INC.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68-02-4229
August 1988
Volume 2, APPENDICES
EPA Task Manager: Margaret Conomos
EPA Project Officer: Philip E Robinson
Prepared for:
Design and Development Branch
Exposure Evaluation Division (TS-798)
Office of Toxic Substances
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
-------
08/01/88 DRAFT
TABLE OF CONTENTS
PAGE
APPENDICES
A. Quality Assurance Project Plans A-l
The Washington Consulting Group A-2
PEI Associates A-22
Midwest Research Institute A-42
B. Data Quality Objectives B-l
C. Letters sent with Questionnaire C-l
D. First Phase Questionnaire D-l
E. Data Tables from First Phase Questionnaire . . . E-l
F. In-Plant Questionnaire F-l
G. Data Tables from In-Plant Questionnaire G-l
H. Individual Site Laboratory Analyses H-l
I. Sensitivity Analysis 1-1
-------
DISCLAIMER
This document has been reviewed and approved for publication
by the Office of Toxic Substances, Office of Pesticides and Toxic
Substances, U.S. Environmental Protection Agency. The use of
trade names or commercial products does not constitute Agency
endorsement or recommendation for use.
-------
APPENDIX A
QUALITY ASSURANCE PROJECT PLANS
A-l
-------
August 10, 1986
QUALITY ASSURANCE PROJECT PLAN, Part I
for the
TEXTILE DYE DRUG ROOM STUDY
by
Bradley Schultz
Washington Consulting Group
1625 Eye Street, N.W.
Washington, D.C. 20006
EPA Contract No. 68-02-4229
Work Assignment 5
EPA Task Manager: Margaret G. Conoroos
EPA Project Officer: Philip Robinson
Design and Development Branch
Office of Toxic Substances
Exposure Evaluation Division (TS-798)
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
A-2
-------
QUALITY ASSURANCE PROJECT PLAN, Part I
for the
TEXTILE DYE DRUG ROOM STUDY
EPA Contract No. 68-02-4229
Work Assignment 5
Approval for:
WASHINGTON CONSULTING GROUP
Approval for:
ENVIRONMENTAL PROTECTION AGENCY
^/"ttAl K
Bryan K. Porter
Quality Assurance Officer
Cr. it
Margaret G. Conoraos
Task Manager
Date
Philip Robinson
Project Officer
Date
Eileen Reilly-wiedow
Quality Assurance Officer
Date
A-3
-------
QUALITY ASSURANCE PROJECT PLAN
There are two parts to this quality assurance plan. The first
part is the statistical design and analysis section. Following
this is Part II on the chemical analysis and field collection
procedures. An overview of the study may be found in the data
quality objectives for this study.
Part I Contents
1. FRAME CONSTRUCTION
2. FIRST PHASE QUESTIONNAIRE
3. RESPONSE RATE FOR SECOND PHASE
4. SELECTION OF PLANTS
5. SELECTION OF WORKER WITHIN PLANT
6. PILOT PLANT MONITORING
7. QA VISITS
8. VERIFYING IN-PLANT RECORDING FORMS
9. DATA ENTRY
10. DATA ANALYSIS
APPENDIX: PROJECT PERSONNEL
A-4
-------
QUALITY ASSURANCE PROJECT PLAN
e«- * Quality assurance (QA) is an integral part of the textile d
study. The following describes the components of the QA program
1. FRAME CONSTRUCTION
SjSSr^ Then this list was cross-checked with a list of pla-Hts^-
known to discharge dye-containing effluent. This list was then
supplemented by EPA personnel familiar with the textile dveinn
;
industry representatives to carefully examine
wer«"h»7n«~mlir""'£* ?ize t0 insure tnat no systematic omissions
were being made. Furthermore, this allowed for a detailed
examination of the eligibility of each plant in thil ripidlv
changing industry (i.e., an «i;~.fwi~ _ii_^ . _ _ .. 4»fi«*y
dyes to color textiles).
2- ' FIRST PHASE QUESTIONNAIRE
Although in-plant monitoring of dye exposure levels i
important aspect of the studv. th*> fire* nKaeA _..«.....;
number of
systematic manner from the kniwn make-up of the dve
th the rame'S acc«racy «0"ld have come into-
* —geable
^ ... . „„ ^ . of
3. allows the possibility of stratification of the sample
for in-plant monitoring (the entire frame of 1390 is too
large to accurately classify with reasonable cost)
A-5
-------
"-
3- RESPONSE RATE FOR SECOND PHASE
.
•onitoring from the plant Tangent ^perspective
1.
a telephone contact to a high level company official
2.
appropriate)
A-6
-------
rocedure*101" BU1 DyS°n (Health * HV9iene) describing
a personal contact to the Chief Executive Officer of the
company, from a Chief Executive Officer of another firm
favorable to the study « 'otner nrm
a letter from a plant that Bill Dyson has already
monitored stating that the monitoring team "was
practically invisible"
site
6* fr°m EPA assurin9 the reluctant plant in writina
^f°rcement action win n°t be taken based on the
lik.l?l!I!IS i?"6" J111 beccarefully crafted to maximize the
likelihood of acceptance of monitoring.
a resPonfe/ate of at least 60% is not achieved, the study
*d ^ that "^ (But thiS W0rst "se scenario^
The refusing plants will be characterized to determine if
" "
by BilirDysoneratin9 plant1' the °"-»ite arrangements will be made
4. SELECTION OF PLANTS
;?f 'Sin1390 PlfntS in the U'S- with the Potential for dyeing
'516 ""^ "mlin
° S!r!;*I*,.deSe;[Bl"td t0 be ineli9ible (i.e., the plants
were out of business, or for some other reason did not
really use powder dyes to dye or print textiles, with
mechanical equipment);
o - 81 had responded by the cutoff date, February 18, 1986;
o 90 did not respond to the mailed-out questionnaire, but
were determined to eligible by EPA (CEB), ATMI, or ETAD
(determination coordinated by CEB's George Heath).
™ SeYeral criteria were considered for stratification, but the
S2f fh'1^6 at W" fc2 ir°up the 171 eli9ible plants by whether o?
not the plant responded to the mailed-out questionnaire Althouoh
whether a plant responded to the mailed-out questionnaire was not
considered to be of direct inherent value as a dividing
A-7
-------
characteristic between types of drug rooms, examination of the
responses to the mailed-out questionnaire suggested that it serves
as a proxy for carpet plants, since it was conjectured that a
higher percentage of non-respondents were carpet plants than the
percentage among respondents. Also the act of responding to the
questionnaire is seen to separate into two groups based on the
likelihood of allowing in-plant monitoring (and thus reducing
somewhat the refusal bias by replacing in-plant refusals with
another plant in the same strata).
A completely random sample of 14 respondent eligibles will be
drawn from the list of 81 respondent eligibles and 16 from the list
of 90 (for a total of 30 plants for in-plant monitoring) in order
to represent plants from each strata with the number of plants
proportional to the total number of plants in each strata. If any
refuse to allow dye level monitoring, they will be replaced by a
plant from the same list that the refusing plant was on.
5. SELECTION OF WORKER WITHIN PLANT
One weigher from each plant will be selected. First, the
shift of the worker will be selected, based upon the facility's (or
plant manager's) seven digit telephone number (obtained from the
first phase questionnaire) as follows:
No of shifts Middle 3 digits of Shift to be
operated • telephone number observed
1 000-999 1st
2 000-499 1st
500-999 2nd
3 000-333 1st
334-666 2nd
667-999 3rd
This will ensure a random selection of shifts for observation and
will provide a mechanism for knowing the shift prior to the visit.
Although not exact, it will allow for a nearly random selection of
worker within plant.
The employee to be monitored will be selected within the
chosen shift as follows:
1. collect last three digits of social security number (SSN)
for all weighers from all dye areas
2. select worker with last three SSN digits closest to 500
A-8
-------
6. PILOT PLANT MONITORING
Methods for in-plant monitoring, recording observations, and
making the chemical analysis, will be pre-tested at a pilot plant
(i.e., an actual plant with textile dyeing operations). *
Of particular note are the testing of:
the ability of the two member team to collect all of the
information, particularly the number of weighings and the
mass of dye measured
the chemical analysis method
7. QA VISITS
Three QA visits will be made during the on-site visits, one at
the beginning, one near the middle, and one near the end of the
site visits. A report will be made on the quality of the data
collected based on information collected as shown in Appendix B.
(At the beginning of the study, corrections may be made to
procedures.) This will also allow the task team to become
intimately acquainted with the usefulness and limitations of the
collected data. Two site visits will be conducted by the Design
and Development Branch of OTS (one of these by their contractor,
The Washington Consulting Group (WCG)) and one by the Field Studies
Branch of OTS. Besides enhancing understanding, the site visits
will check the procedure for selecting the shift and weigher chosen
(and thus also the drug room chosen), the determination of the
number of weighers, and all other data collected at the site ~ but
with particular attention paid to those items which are recorded in
prespecified categories. Part II, sections 7 and 10 discuss
chemical audits in detail.
8. VERIFYING IN-PLANT RECORDING FORMS
In-plant data collection forms will be compared by the two on-
site industrial hygienists for accuracy of the information.
9. DATA PROCESSING AND ENTRY
ATMI mailed out the first phase questionnaire. After return
to ATMI the coded, unidentified questionnaires were forwarded to
EPA and then to WCG. All questionnaire/responses were computerized
by WCG and double-entered to reduce keypunch e.rrors to a minimum.
George Heath, of EPA's Chemical Engineering Branch, who is familiar
with the dyeing industry and its operations, may provide a list of
obviously incorrect questionnaire answers where there was a
misunderstanding or other error on the part of the respondent (if
corrections appropriate). The list of changes would be provided to
the study partners (EPA. and industry) along with a brief
explanation, and this list and explanation would be provided with
the summary table results. At the end of the study, all response
forms will be returned to ATMI and destroyed.
A-9
-------
fiPiH n 5 ^6 feP°nd-Phase data handling will be done in the
field and at the laboratory (discussed in part II of QA plan)
fh H I 6 SJf6 V^Sit 3nd the verification of collected information
the data collection protocol will be split into two parts: thJJSJ'
monitoring cartridges and accompanying information, and the *
?o SfrL Ti resultsl The observational results will be fo'rwarded
Geo?ge HeIth?°n0mOS ° th6 DeSi9" 3nd Devel°Pment Branch, and
of^hTfo n?fntWi^ ,the" enter the 'ollowing information "or each
of the 30 plants, into a computer data base for statistical
analysis, and presentation of summary tables:
1. 2-digit plant identification number
2. date (month, day, year)
22.
23.
°f W6i9herS W0rkin9 at ?lant (all work areas, all
5-
observed shift
^
?* "ave"gef ^^ dyei"9 machi"es Cunning during observed shift
8.
in' ™bep °f ?ther workers Corking in observed drug room shift
10. age of worker at time of site visit (number of Sears)
?rofeynrsrCe °f m°nit°red WOrker handlin9 *****
*' °bSerVed facilitV ** Ponder dye handler (number
14* «hfJ!!er °r n°! wef9her wore d"st «ask during shift (yes/no)
15* wh^h °r n°l wef9her ate in drug room during shift lyes/no)
Tyes/no)°r "° W619her smoked in dru« room du^"9 shift
18* SumSfr 25 ^ wei?hed fay worker ob'served during shift
rf raonitored ^as in drug rlom (nearest
-erage
21 ' hSur" °f hOUFS personal »onitor running (nearest tenth of
gravimetric weight of all dust on personal monitor cartridge
gravimetric weight of all dust on Srea sampler cartridge
A-10
-------
10. DATA ANALYSIS
There will be two major components to the data analysis:
1- and^g^e^herl distribution °f -V ^vels in plants
2. examining the correlation of dye level with other factors
Other items of interest are discussed in section 10.4 - 10.5.
10-1 DISTRIBUTION OF DYE LEVELS IN PLANTS
chcriMd. ' ist"bution of plant dye levels will be
10'2 ESTIMATE OF AVERAGE WORKER EXPOSURE
A-ll
-------
KUrini a JyPical 24 hour Period. For example, statements
could be made about the 85th percentile of worke? exposure by
Thi! ii"?*P lant •»?iMte» by the number of weighers in each plant.
This would result in a statement such as "It is estimated that 8si
hLr JSfJJV'J eXP°Sed t0 leV6lS lower than xxx m*/m1 during an 8
nSrl. ?^f '% ?te that this differs f'°n> the estimate of 85th
KISS J6 °fKplant Jevels (objective 1) which results in a
!« !J ? f"fh aS!- Ifc is estimated that the average exposure in
tlL LiSfiJ%dj;ein9 and Printin9 Plants is less than zL mg/m*?
time-weighted 8-hour average" (xxx and zzz determined from study).
from
10.3 CORRELATION OF DYE LEVEL WITH OTHER FACTORS
10. 3A Primary focus
The exposure level (mg/m3) will be examined for
correlation with:
1. mass of weighing during shift by worker of interest
2. number of weighings during shift by worker of interest
J. combination of the two above factors.
If such a correlation exists, a functional relationship will
be explored between the dye level and other factors.
10-3B Secondary measurements to examine association with
dye level --- •
___ ?h?-rate.°£ exposure (mg/m3/hr) will also be examined for
These are?" several °ther variables of secondary importance.
1. Production volume of textiles (pounds per year), from
mailed-out questionnaire
2. Management of dye house (vertical, commission or both),
from mailed-out questionnaire
3. Management of dye house (public or private), from mailed-
out questionnaire
4. Color index class of dyes used, for any dye used during
observed shift by monitored weigher (acid,
basic/cat ionic, reactive, direct, disperse, other), from
site visit log as classified by Chemical Engineering
Branch
5. Total number of dyeing and printing machines serviced by
monitored weigher (average of beginning and end of shift
numbers), from site visit questionnaire
6. Fiber type dyed or printed (acrylic/modacrylic,
rayon/cotton, nylon, polyester, other), from mailed-out
questionnaire
A-12
-------
o£
10.4 AMOUNT OF DYE WEIGHED OUT
10.5 SUMMARY TABLES
10.5.1. On-site questionnaire
lB region °f textile dyein9 Plants in "<* EPA geographical
& •
3.
4.
5.
6.
7.
8.
9.
10.
Number of plants by pounds of dye weighed during shift
Number of plants by number of djes weighed during shift
Number of plants by number of dje weighings during shift
Number of workers by amount of time in drGg room 9
Number of workers that used dust mask during site visit
Su±r °J W°r!!erS that USed r«pirator during site vfsit
Number of workers that smoked in drug room a?ea during site
Number of workers that ate in drug room area during site visit
A-13
-------
10.5.2 Mailed-out questionnaire
As an appendix in the final report, the following variables
will be tabulated from the first-phase, mailed-out questionnaire:
*11. Number of textile dyeing plants in each EPA geographical
region * *
12. Number of plants by number of dyeing or printing operations
within the company that owns the selected plant
§13. Number of plants by management of house (vertical, commission.
or both)
§14. Number of plants by management of house (public, or private)
§15. Distribution of plants by product volume
*?S* SUmw6r °! pjants fay Product line (carpet, yarn, fabric, other)
§17. Number of plants by type of dyeing or printing equipment
ava (batch, semi-continuous/continuous, printing)
§18. Number of plants by fiber dyed or printed (acrylic/raodacrylic
rayon/cotton, nylon, polyester, other)
§19. Number of plants by color index class of powder dye (acid,
basic/cationic, reactive, direct, disperse, other)
20. Number of plants by number of dyes weighed per 24 hours (less
than 10, 10 to 20, over 20)
&21. Number of plants by pounds of dye used per 24 hours (less than
50, 50 to 200, over 200)
&22. Number of plants by number of powder dye weighings per 24
hours (less than 50, 50 to 500, over 500)
23. Number of plants by number of dye weighing rooms ( 1 room. 2
or more rooms)
24. Number of plants by number of worker shifts per 24 hours (1,
2, 3)
25. Number of plants by number of operating days per week
(1 to 4, 5, 6 or 7)
26. Number of plants by number of employees exposed to powder dyes
(1, 2, 3, 4 or more)
Notes
* This is also tabulated from site visit data
I Similar information is also collected in the on-site
questionnaire and is used in the secondary correlation
analysis (see Appendix A)
§ A portion of the data on this variable will be used in the
secondary correlation analysis (see Appendix A)
& Similar information also collected in the on-site
questionnaire and is used in the primary correlation analysis
A-14
-------
10.6 AREA SAMPLER RESULTS
Although the area sampler measurements will be chemically
analyzed for each plant, the results will be used solely for post-
study exploratory work and possibly for quality assurance purposes
(if a strong correlation is found between personal samplers-&nd
area sampler results). The area samplers may provide a useful
quality assurance role at the data analysis phase. If the personal
monitor result is suspect at one plant for some reason, the area
sampler measurement provides a rough cross-check for such a suspect
value. As in any study, it is hoped that no such incidents will
take place.
A-15
-------
APPENDIX; PROJECT PERSONNEL
Bradley Schultz is the WCG work assignment leader. He will be
the statistician involved in the design and analysis of the study
David Cox is the overall Project Director for the WCG contract with
EPA. Doug Marder and Keith Johnston will set up and oversee the
creation and use of the data bases. Terri Stiteler will manage the
•I! ^S6^ d" C001[dinate the data entry. Credentials are on file
with the Design and Development Branch of EPA.
A-16
-------
APPENDIX B
TEXTILE DYE/DPUG ROOM EXPOSURE STUDY
QUALITY ASSURANCE AUDIT FOR A PLANT VISIT
Section I. Basic Audit Information
A. Auditor Information
1. Name(s)/Affiliation:
2. Date of Audit: , 19
B. Textile Plant Information
Plant I.D.
C. Industrial Hygienists:
1. Health and Hygiene, Inc.
2. PEI
A-17
-------
Section II. Sampling Design
1. Was the selection of the shift and/or the weigher done on a
•»
random basis? Yes No
A-18
-------
Section III Air Monitoring Methods
1. Were all air sampling instruments calibrated
accurately prior to field use? (Note means of
verification). Yes No
2. v;ere the personal samplers operating at a flow rate between
2 and 2.5 liters/minute? Yes No
3. Did anything occur that might interfere with the airflow on
the personal sampler (i.e tubing became twisted)?
Yes No
4. Were the area samplers operating at a flow rate between 5
and 8 liters/minute? Yes No
5. Was the location of the area samplers appropriate?
Yes No
6. Were filter blanks taken into the field? Yes No
7. Did splashing of liquids occur onto the filter?
Yes No
A-19
-------
Section IV Drug Room Observations
1. Was each container of bulk dye taken labelled and
appropriately identified on the corresponding form? *
Yes No
2. Were appropriate methods used to collect bulk dye samples,
i.e. non-obtrusive and non-dust-generating? Yes No
3. Were all entries and exits into and out of the drug room by
the weighes recorded? Yes No
4. Were all weighings recorded with name of the dye noted and
corresponding to the name of the bulk dye sampled?
Yes No
5. Was a validation of all dyes recorded conducted between the
two industrial hygienists? Yes No
A-20
-------
Section V Qualitative Performance of the Field Visit
1. Was a positive rapport evident between nnnagenent and the
visiting industrial hygienists? Yes No •»
2. v;as management well-informed about the objectives of the
study and fully co-operative? Yes No
3. Was the weigher who was wearing the sampling device
informed about the objectives of the study and fully
cooperative? Yes No
4. Were the weigher's work activities altered or interrupted
by the visiting industrial hygienists? Yes No
A-21
-------
Ecological ana To».coiogicai Associahon ol me Dyesiulls Manu(aciunng
>-'S j°L*f f.j CCW'"EL OF ET-D
August 12. 1986
Mr. George Heath
TS-779
U.S. Environmental Protection Agency
Office of Pesticides &
Toxic Substances
401 M Street. SW
Washington. DC 20460
Re: ATOI/EPA/EITO Exposure Project
Dear George:
t «.. X cni:lose a copy of Bill Dyson's revised quality assurance
for tJus project. This seems to incorporate the Sarious c^S^
July 30' U86 ** discussed the proposal that
be conducted by MRI rather than by BurUngto
Hygiene laboratory. It ves agreed that
Chan9e* Hc^verhaving disc r m
sh * c^veravng isc r
Dyson such a change does not seem to offer significant logistical benefits
S'ittffJSS™' I S— - ^- ^ SS2
as
Clarke
Jtive Secretary
EAC/bss
Encl
Fef. US/&P.3.1.
1330 Connecticut Avenue. N W. Washington. D C 20036 • (202) 659-0060 • Telex 6971018 SOC ETAD
A-22
-------
DRAFT
Section: 1.0
Revision: 1
Date: July 14, 1986
Page: 1 of 1
SECTION 1.0
EPA/ETAD/ATMI
DYE EXPOSURE STUDY
QUALITY ASSURANCE PROGRAM
FOR
FIELD SAMPLE AND DATA COLLECTION
A-23
-------
Sect;on: 2.0
Revision: 1
Date: July 14, 1986
Page:. 1 of 1
SECTION 2.0
TABLE OF CONTENT
Section Heading
1.0 Title Page
2.0 Table of Content
3.0 Project Description
4.0 Facilities, Equipment
Consumables, and Services
5.0 Sample-and-Data Generation
6.0 Data Processing
7.0 Data Quality Assessment
8.0 Corrective Action
9.0 Documentation and Reporting
10.JO Personnel and Management
Pages Revision Date
1 1 07/14/86
1 1 07/14/86
3 1 07/14/86
3
6
1
1
1
1
1
1
1
1
1
1
1
1
07/14/86
07/14/86
07/14/86
07/14/86
07/14/86
07/14/86
07/14/86
A-24
-------
Sec t. j en: 3 . r
Pevjsjon: 1
Date: July 14, 1986
Page: 1 of 3
SECTION 3.0
PROJECT DESCRIPTION
The American Textile Manufacturers Institute (ATMI), the U.S.
Operating Committee of the Ecological and Toxicologica1
Association of the Dyestuffs Manufacturing Industry (ETAD), and
the U.S. Environmental Protection Agency (EPA) are jointly
sponsoring a study to assess the potential exposure of dye
weighers in textile drug rooms to airborne dye dust. Data from
this study will be used to estimate worker exposure as new dyes
are proposed for introduction into commerce and in addressing
concerns on existing products.
The study is divided into three parts. The first is the
selection of textile facilities to be visited. This will be done
on a random basis from the total universe of textile dyeing and
printing facilities in the U.S. by the Washington Research Group
under contract to the EPA with assistance from ETAD and ATMI.
The second is the collection of data and samples at the 30
facilities selected. This will be done by Health & Hygiene, Inc.
as a contractor to ETAD and ATMI. and is the subject of this
quality assurance plan. The third is the analyses of samples
collected at the facilities to determine the dye content of the
airborne dust in the drug rooms. This will be done by Midwest
Research Institute as a contractor to the EPA and is the subject
of a separate quality assurance document.
The field sample and data collection part of this study is
comprised of the following activities:
3.1 Preliminary Arrangements
Agreement by a selected facility to participate in the
study will be obtained by ATMI with assistance from
ETAD and others. Once consent has been given,
telephone contact will be made by Health & Hygiene to
schedule a visit. Preliminary information necessary to
prepare for the visit such as the number of shifts of
drug room operation and the approximate number of dyes
weighed per shift, will be obtained during this contact
(Attachment 3-1). Other activities including obtaining
necessary supplies, calibrating air sampling pumps, and
preweighing sample filters and blanks will be done
prior to the visit.
A-25
-------
:.:.: 3.G
Revision: 1
Date: July 14, 1986
Page: 2 of 3
3.2 Field Visits
Visits to the selected facilities will be made by a two
person team of industrial hygienists, one each from
Health & Hygiene, Inc. and PEI Associates. Inc. The
team will arrive at the facility at least four hours
prior to the start of the shift selected for
sampling/observation to obtain information about the
facility, determine how to obtain dye weighing data
most efficiently, and take a brief familiarization
tour. During the selected shift, personal and area air
samples will be taken; bulk samples of dyes weighed
during the shift collected; temperature, barometric
pressure, and relative humidity measurements made; the
time the monitored employee spends in the drug room
recorded; data on the number and quantity pf dyes
weighed obtained; and observations about conditions and
controls in the drug room made. Upon completion of the
sampling and observations a closing conference will be
held with management.
3.3 Gravimetric Determinations
Total potential dust exposure of the employee during
the monitoring period will be determined by reweighing
the air sampling filters collected at each facility.
After weighing, the filters and bulk dye samples will
be sent to Midwest Research Institute (MRI) for
analytical determination of the dye content on the
filters. The gravimetric determination of total dust
will provide an upper bound against which the
nonspecific analysis by MRI can be compared.
A-26
-------
: i i •_ :i: .-• . 0
Revision: 1
Date: July 14. 1986
Page: 3 of 3
ATTACHMENT 3-1
PRE-VISIT TELEPHONE CONTACT
Facility:
Address:
Contact:
Telephone No.
Scheduled Visit Date:
Number Of Shifts Drug Room Operated:
Shift to be observed/sampled:
Comments:
Approximate number dyes weighed/shift:
Approximate cumber dyes weighings/shift:
Local accommodations:
A-27
-------
SecLicn: 4.0
Revision: 1
Date: July 14, 1986
Page: 1 of 3
SECTION 4.0
FACILITIES, EQUIPMENT, CONSUMABLES, AND SERVICES
4.1 Facilities
Filter weighing and reweighing will be done at Burlington
Industries' Industrial Hygiene Laboratory. Doyle Street.
Greensboro. North Carolina. This laboratory has a constant
temperature-humidity room in which the filters and blanks
can be equilibrated prior to each weighing.
Sample pump calibrations will be performed in the industrial
hygiene laboratory at Health & Hygiene. Inc.
4.2 Equipment
Equipment which will be used on this project includes:
• Mettler ME 30 microbalance, capable of weighing to the
nearest microgram
• Gilian HFS 113 air sampling pumps with timers
• Cast Model 1531 vacuum pumps with critical orifices
• GCA/Precision Scientific wet test meter
• Buck Model M-5 mini-calibrator
• Bacharach Sling Psychrometer
4.2.1 Calibration
The Mettler ME 30 microbalance is checked prior to
each set of weighings with an internal 100 mg
weight. Zero checks are made periodically during
the weighings to assure that drift is not
occurring. If the balance cannot be calibrated or
zeroed a Mettler service representative will be
called.
Flow rates for the Gilian sampling pumps will be
calibrated in the laboratory prior to field visits
with an SKC 311-100 soap film calibrator. Prior
to site visits while in the field the Buck mini-
calibrator will be used. Further, the
A-28
-------
Stc-•_._:•: 4.0
Revision: 1
Date: July 14. 1986
Page: 2 of 3
rotameter setting will be noted during calibration
and checked periodically during sampling. Should
low flows be noted during sampling, a post-
sampling flow rate will be determined with the
mini-calibrator and an average value used.
Flow rates for the Cast vacuum pumps with critical
orifices will be determined prior to field visits
with the wet test meter.
4.2.2 Maintenance
Maintenance of equipment used in this project will
be done according to the manufacturer's
specifications. The Mettler microbalance has just
been serviced and is on a yearly schedule. All
other equipment is serviced as needed.
4.3 Consumables
Consumables supplies which will be used during this project
include:
• Gelman Vinyl Metricel VM-1 filters with support pads
• Gelman 4339 3-piece cassettes
. UV Light Absorbing Plastic Bottles (for bulk sainples)
• Plastic spoons
• Plastic bags
4.4 Services
Filter weighing will be done at Burlington Industries'
Industrial Hygiene Laboratory. Services to be provided by
this laboratory include:
• Equilibrate, preweigh, and place filters in numbered
cassettes prior to field visits, including
• field blanks
• Equilibrate and reweigh filters and blanks after
sampling
A-29
-------
Section: 4.0
Revision: 1
Date: July 14, 1986
Page: 3 of 3
Place samples in unused cassettes
Record results in a permanent laboratory notebook
Send samples, cassettes, field blanks, filter blanks
and recorded results of weighings to MRI
A-30
-------
St:. !_•„:!: F. . G
Revision: 1
Date: July 14, 1986
Page: 1 of 6
SECTION 5.0
SAMPLE AND DATA GENERATION
5.1 Employee Selection
For each of the thirty (30) facilities visited, only one dye
weigher will be selected for exposure monitoring. This will
be done in a random fashion. the shift to be observed and
sampled will be selected based on the middle three digits of
the facility's (or facility contact's) seven digit telephone
number as follows:
No. Shifts Middle 3 digits Shift to be
Operated of telephone number observed/sampled
1 001-999 1st
2 001-500 1st
501-999 2nd
3 001-333 1st
334-666 2nd
667-999 3rd
This will ensure a random selection of the shift for
sampling and will provide a mechanism for knowing the shift
prior to the visit. If contact with the facility indicates
that the shift selected in the above fashion is unreasonable
due to the dye weighing activities being conducted, then an
alternative selection will be made based on professional
judgment.
Where the facility has more than one dye weigher on the
selected shift, the one selected for monitoring will be the
dye weigher whose last three Social Security Number digits
are closest to 500. Assistants and helpers will not be
considered for inclusion. Selection of the employee will be
done in the opening conference with management. The method
of selection will be documented for both shift and weigher
if different from normal procedure (page 2 of survey form.
A-31
-------
Section: 5.0
Revision: 1
Date: July 14, 1986
Page: 2 of 6
5.2 Air Sample Collection
At each facility visited, a total of four air samples will
be collected, two worn by the dye weigher selected for
monitoring and two at stationary locations in the drug room.
Air will be drawn through pre-weighed 37 mm polyvinyl
chloride filters (Gelman VM-1) in three-piece cassettes.
Personal samples will be taken using portable, battery
operated pumps worn by the employee as he performs his work.
Flow rates of approximately 2.0 L/min and a sampling
duration of seven or more hours, including breaks, will be
used. Stationary area samples will be obtained using
electrical vacuum pumps operated at approximately 7.4 L/min
for roughly the same duration as the personal samples.
Sampling will-be -done in an "open-face" configuration.
After collection, the samples will be recapped, returned to
the laboratory, equilibrated, reweighed to the nearest
microgram, and submitted to MRI for determination of dye
content.
5.2.1. Sample Identification
As the filters are preweighed and placed into
cassettes, a unique identifying number will be
written on the bottom section of the cassette with
an indelible marker. This will serve as the
sample identifier on all documentation of the
visit and subsequent analysis. When the filters
are reweighed after collection and placed into new
cassettes, the bottom section of the new cassettes
will be identified by the same number as the
sample plus a prime (•) mark to distinguish it
from the original cassette. Both will be
submitted to MRI.
5.2.2 Duplicate Personal Samples
Both personal samples at a facility will be
collected from the same dye weigher.
5.2.3. Air Flow Checks
The personal sampling pumps used have electronic
feedback systems which maintain set flow rates
even as dust on the filter increases resistance.
These pumps also have rotameters. The setting of
the rotameter will be noted during calibration and
A-32
-------
Sec Lion: fi.J
Revision: 1
Date: July 14, 1986
Page: 3 of 6
observed two to three times during sampling to
ensure that no change has occurred. If it appears
that the air flow rate has changed, a post-
sampling flow calibration check will be made using
the mini-Buck calibrator. Should significant
variation be found the average of the pre- and
post-sampling calibrations will be used as the
flow rate for dust level calculations.
Air flow rates for the two area samplers will be
maintained with critical orifices. These are
generally quite steady. A rotameter will be used
to check air flow periodically during the sampling
period. Should significant variation be found, a
post-sampling calibration will be made and the
average value used as the flow rate.
5.3 Data Collection
Information on the time the dye weigher being observed
actually spends in the drug room, the number of dye
weighings made, and the total quantity of each dye weighed
during the sampling period will be obtained. Observations
will be made of ventilation in the drug room, the use of
personal protective equipment, work practices being used,
and cleanliness of the area. In addition, limited
demographic data will be obtained from the dye weigher being
monitored/observed and temperature, barometric pressure, and
relative humidity in the drug room will be recorded. The
use of any particularly dusty dyes will be recorded.
5.3.1 Exposure Duration
Dye weighers are typically quite mobile. They
move freely in and out of the drug room. Since it
is assumed that the majority of exposure occurs
while he is in the drug room, this parameter will
be measured.
The boundary of the drug room will be established
by agreement between the two site visitors. A
rough sketch of the dye weighing area will be made
indicating the boundary chosen. The actual time
the dye weigher being observed spends in the drug
room will be measured by recording the time, to
A-33
-------
Section: 5.C
Revision: 1
Date: July 14, 1986
Page: 4 of 6
the nearest minute, that he enters and exits the
area. The total potential exposure time will be
the sun of the periods he spends in the drug room.
5.3.2 Weighings of Solid Materials
All weighings of solid materials made during the
sampling period will be observed to determine
three important parameters - the total number of
individual weighings made, the total number of
individual solids weighed, and the total mass or
quantity of each solid .weighed. Whenever
possible, batch tickets used by the dye weigher
will be obtained and the weighing data recorded
from these. If batch tickets are not available-^-
the data will be obtained by direct observation
and, if necessary, questioning the dye weigher.
The method for obtaining these data will be
discussed with management in the preliminary
conference. Their suggestions as to the most
efficient way to obtain the data in their facility
will be considered and used where possible.
5.3.3 General Observations
During the sampling period, general observations
will be recorded about the type of personal
protective equipment used by the dye weigher, work
practices used, the cleanliness of the drug room,
and engineering controls such as ventilation used
to reduce dust exposure. Whether or not the dye
weigher smokes will also be recorded.
5.3.4 Dye Weioher Interview
During a break period, the dye weigher will be
asked several questions to determine his age and
how many years he has been handling dyes, both at
the facility visited and elsewhere.•
5-3.5 Temperature/Relative Humidity
Temperature, barometric pressure, and relative
humidity will be measured intermittently during
the sampling period.
A-34
-------
Section: 5.0
Revision: 1
Date: July 14, 1986
Page: 5 of 6
5.4 Bulk Dye Samples
For each individual dye weighed during the sampling period,
a bulk sample of approximately one ounce will be obtained.
These samples will be taken from the original dye drums used
by the weigher. Disposable plastic spoons will be used to
avoid contamination.
The bulk dye sample bottles will be labeled sequentially
using the labels provided by MRI. Corresponding labels (of
the same number) will be placed on the data sheet where the
full name of the dye. its lot number, if possible, and its
supplier will be recorded along with the name used on the
batch ticket or weighing record. Samples will also be taken
of those solids wheighed by the weigher which may interfere
with the dye analysis (e.g. colored materials or chemicals
which may react with dyes).
5.5 Filter Weighings
Air sampling cassettes will be returned to Burlington
Industries' Industrial Hygiene Laboratory. They will be
equilibrated overnight in the constant temperature and
humidity area prior to being reweighed to the nearest
microgram. Ten field blanks - preweighed filters in
cassettes through which no air has been drawn which have
been handled and transported with the sample filters - will
be equilibrated and weighed at the same time. The average
weight change of these ten blanks will be used as the blank
correction values in calculating dust exposure levels.
Significant problems in filter weighing should be detected
through the use of these blanks.
Note: The ten field blanks may apply to air samples from
more than one facility if two or more are visited in the
same week.
5.6 Shipping and Handling
After collection and reweighing, all materials will be
shipped to MRI. This includes bulk samples, air sample
filters, and data collection forms.
Bulk sample bottles will be placed in plastic bags in groups
of four. These will be shipped to MRI separately from the
air samples to avoid the potential for contamination. To
minimize handling, bulk samples will be sent directly from
Health 6 Hygiene.
A-35
-------
Section: 5.0
Revision: 1
Date: July 14. 1986
Page: 6 of 6
Air samples will be hand carried to Burlington Industries'
Industrial Hygiene Laboratory. After reweighing. they will
be shipped directly to MRI. For each facility, these
shipments will include the four air sample filters in new
cassettes, four empty field cassettes, two field blanks, and
ten unused filters from the same lot.
Data collection forms, minus the identifying cover sheet,
will be hand delivered along with the air samples to
Burlington's laboratory. After post-sampling weights are
recorded, the forms will be sent along with the samples to
MRI. The facility identification sheet will be mailed to
ATMI from Health & Hygiene.
A-36
-------
Set. i ;o:.: b.v.
Revision: 1
Date: July 14. 1986
Page: 1 of 1
SECTION 6.0
DATA PROCESSING
6.1 Collection
All data will be collected manually. Recording will be done
legibly in permanent ink on worksheets. Each person
involved in recording data will sign and date the worksheet.
Corrections will be initialed.
6.2 Data Reduction
All sample manipulations will be clearly documented.
Standard data reduction techniques will be used.
6.3 Data Validation
The data validation process will include:
• Air flow rate checks
• Timing checks with second watch
* Checking calculations
• Comparisons with original batch tickets
• Reviews for internal consistency by site visitors
• Use of field blanks
The site visitor from Health fc Hygiene will be responsible
for assuring data validity.
6.4 Transfer
Original recording sheets will be included with field visit
documentation to allow checking of data transfers at a later
date.
A-37
-------
Seel:en: 7.0
Revision: 1
Date: July 14, 1986
Page: 1 of 1
SECTION 7.0
DATA QUALITY ASSESSMENT
7.1 Filter Weighings
The microbalance used for filter weighings is capable of +1
microgram precision and, since net weight change is being
measured, accuracy. However, zero drift is slightly greater
than this, approximately +3 micrograms. Zero will be reset
after every third weighing.
Greater variability is caused by moisture collection on the
filters. A constant temperature and relative humidity room
is used to equilibrate the filters before weighings to
reduce this variability. Ten blank filters through which no
air has been drawn are weighed and reweighed with the sample
filters. The average weight change of these blanks is used
as a blank correction in calculating dust levels. This
correction is generally less than +30 micrograms.
7.2 Traceabilitv of Samples and Data
All air and bulk samples will have unique identification
numbers. All data collected on the samples will be related
to these numbers. Simplified traceability logs will be
completed and signed when samples are transferred to
Burlington's laboratory and to MRI.
7.3 Completeness
The Health & Hygiene site visitor will review all data
collected prior to leaving the facility to assure
completeness. This will include a verification that the
record of dyes weighed/sampled is consistent.
A-38
-------
Section: f.0
Revision: 1
Date: July 14, 1986
Page: 1 of 1
SECTION 8.0
CORRECTIVE ACTION
The Health & Hygiene site visitor has primary responsibility for
taking corrective actions as necessary. Examples of problems
which might be encountered and possible corrective actions are as
follows:
Personal sampling pump stops - Record time from built-in
timer, submit sample, make second sample primary
Flow rate variation during sample - average pre- and post-
sampling rates
• Large blank filter weight variation - check balance, examine
filters for loss of material or contamination from backup pad.
• Failure to obtain bulk dye sample - Contact facility for
assistance in obtaining
• Incomplete dye weighing data - Obtain original batch tickets
from facility if possible
• pater spray or other inadvertent contamination of air sampling
filter - use second sample if not contaminated, conduct
sampling again
A-39
-------
S-rCliG:-,: 9.0
Revision: 0
Date: June 9, 1986
Page: 1 of 1
SECTION 9.0
DOCUMENTATION AND REPORTING
9.1 Documentation
Field sample and data collection will be documented in
permanent ink on the forms provided by the EPA. The data
related to sample collection will be compiled by the Health
& Hygiene site visitor and Burlington lab personnel. Other
data will be completed by the PEI, Inc. site visitor. Any
corrections will be marked through and initialed. Raw data
on air flow calibrations and filter weighings for which no
space is provided on the EPA forms will be recorded on
standard forms from Health & Hygiene and accompany the
documentation for a facility visit.
9.2 Transmittal to MRI
MRI is acting as the central repository for all data related
to this study. Data collected prior to and during facility
visits will be submitted to MRI along with the samples for
that facility by both Health & Hygiene and PEI. The
identification sheet for each facility will be sent
separately to ATMI by Health & Hygiene.
9.3 Reports
Trip reports will be prepared by the PEI. Inc. site visitor
for each facility visited. Upon completion of the study, a
composite report of dye dust exposure in textile drug rooms
from an industrial hygiene point of view will be prepared
jointly"by PEI, Inc. and Health & Hygiene, Inc.
A-40
-------
SC-CL :,;::: 10.0
Revision: 1
Date: July 14, 1986
Page: 1 of 1
SECTION 10.0
PERSONNEL AND MANAGEMENT
10.1 Health & Hygiene Personnel
Dr. William L. Dyson. CIH, Vice President will have primary
management responsibility for this project at Health &
Hygiene. Other personnel who may participate in the study
are Ronald Hill. CIH and David S. Davis, both industrial
hygienists. The Burlington Industries Industrial Hygiene
Laboratory person for the project is Sharon Lonon. She has
more than seven years experience with filter weighing for
cotton dust and other sampling at Burlington.
10.2 PEI. Inc. Personnel
Personnel from PEI who may participate on this project are
Leslie J. Ungers, CIH and Robert W. Willson, CIH.'
10.3 Project Coordination
Scheduling visits to the textile facilities will be done by
Health & Hygiene. Conducting the preliminary management
conference and coordinating sample and data collection at
the facility are the responsibility of the Health & Hygiene
site visitor.
A-41
-------
DYES - ANALYTICAL METHODOLOGY DEVELOPMENT AND
j ANALYSIS OF FIELD SAMPLES
i
DRAFT QUALITY ASSURANCE PROJECT PLAN
for the
Office of Toxic Substances
EPA Prime Contract No. 68-02-4252
Work Assignment No. 56
MRI Project No. 8856-A(01)
For
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch, TS-798
Washington, D.C. 20460
Attn: Mr. Richard Kent
A-42
-------
DYES - ANALYTICAL METHODOLOGY DEVELOPMENT AND
ANALYSIS OF FIELD SAMPLES
DRAFT QUALITY ASSURANCE PROJECT PLAN
for the
Office of Toxic Substances
EPA Prime Contract No. 68-02-4252
Work Assignment No. 56
MRI Project No. 8856-A(01)
For
U.S. Environmental Protection Agency
Office of Toxic Substances
Field Studies Branch, TS-798
Washington, D.C. 20460
Attn: Mr. Richard Kent
MIDWEST RESEARCH INSTITUTE 425 VOLKER BOULEVARD. KANSAS CITY. MISSOURI 64110 • 816 753-7630
A-43
-------
Section No.: 1.0
Revision No.: 3
Date: April 30, 1987
Page 1 of 1
SECTION 1.0
DYES - ANALYTICAL METHODOLOGY DEVELOPMENT AND
ANALYSIS OF FIELD SAMPLES
Draft Quality Assurance Project Plan
EPA Contract No. 68-02-4252
Work Assignment No. 56
Approval for:
MIDWEST RESEARCH INSTITUTE
Program Manager
Approval for:
ENVIRONMENTAL PROTECTION AGENCY
Joseph J. Breen
Project Officer
BatT
oTsinger-—J Date^
Quality Assurance Coordinator
tiieen ReiIly-WiedowDate
Quality Assurance Officer
A-44
-------
Section No.: 2.0
Revision No.: 3
Date: April 30, 1987
Page 1 of 1
SECTION 2.0
TABLE OF CONTENTS
Section
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
Appendix
Appendix
Heading
Title Page
Table of Contents
Project Description
Project Organization and Management
Personnel Qualifications
Facilities, Equipment, Consumables,
and Services
Data Generation
Data Processing
Data Quality Assessment
Corrective Action
Documentation and Reporting
A - Standard Operating Procedure for
the Cary 219 Spectrophotometer
B - Analytical Protocol
Pages
1
1
2
4
1
3
8
2
3
2
2
Checking the
Revision
3
3
3
2
2
3
3
3
2
2
2
Calibration
Date
4/30/87
4/30/87
4/30/87
4/30/87
4/30/87
4/30/87
4/30/87
4/30/87
4/30/87
4/30/86
4/30/87
of
List of Plan Holders:
Midwest Research Institute:
J. Spigarelli, J. Going, P. Constant, J. Hosenfeld, J. Balsinger,
C. Green, D. Harbin, J. Long, R. Ayling, R. Rembecki
Environmental Protection Agency:
J. Breen, E. Reilly-Wiedow, R. Kent
A-45
-------
Section No.: 3.0
Revision No.: 3
Date: April 30, 1987
Page 1 of 2
SECTION 3.0
PROJECT DESCRIPTION
The Environmental Protection Agency (EPA) has initiated a joint study with the
Ecological and Toxicological Association of Oyestuffs Manufacturing Industry
(ETAD) and the American Textile Manufacturers Institute (ATMI) to assess the
exposure of textile plant dye weighers to airborne dye particles which are
present in so-called "drug rooms" in the textile facilities. As prime con-
tractor for the Office of Toxic Substances, Midwest Research Institute (MRI)
has been directed to develop the analytical methodology necessary to determine
the total amount of dyes present on air sampling filters.
After extensive discussions between industrial dye chemists and analytical
chemists at MRI, it was concluded that conventional quantitation of individ-
ual dyes on each air filter was not feasible, given the low quantities ex-
pected to be present. Various alternative methods based on measuring phys-
ical properties of dyes as a class of compounds were considered. A method
based on spectrophotometry was deemed to be the most applicable to the analy-
sis of dyes in general, especially at low levels.
The use of quantitation methods that are general (i.e., nonspecific) for a
class of compounds will frequently result in final values which are more un-
certain than those values obtained from a more specific method. The spectro-
photometric approach to the determination of total dyes on an air filter will
result in an estimation of the amount of dyes that are present. This is be-
cause the method assumes that all dyes that are weighed in the drug room are
present on the air filter and appear in amounts proportional to the amount of
each dye handled during the air monitoring period. Adding to the uncertainty
of the value is the inability to determine (in most instances) the suitability'
of using these analytical assumptions in analyzing the actual samples. One
distinct advantage of the spectrophotometric method, however, is that its
accuracy improves with the number of dyes present on the filter, i.e., in-
creasing sample complexity will give better dye estimates.
A means of assessing the approximate uncertainty of the dye estimate can be
found by undertaking a statistical treatment of the absorbance characteristics
of the individual dyes comprising the sample set. In this fashion, probable
errors in the dye estimate can be generated for different dye mixture scenarios
of the sample set. Absorbance spectrum profiles and/or drug room dye utiliza-
tion data can then be employed to either favor or rule out certain dye mixture
scenarios.
The scope of work is comprised of two tasks as described below.
3.1 Subtask A: Analytical Methodology Development
MRI will develop analytical methodology for estimating the amount of
dyes collected on a filter during air sampling in dye drug rooms. The
A-46
-------
Section No.: 3.0
Revision No.: 3
Date: April 30, 1987
Page 2 of 2
experimental lab study will focus on estimation of unknown quantities
of dyes by a visible absorbance method. Various aspects of this dye
estimation method will be investigated. These include determining the
detection limits of various groups of dyes, establishing the uncertainty
of the estimate, devising an efficient filter extraction scheme, and de-
termining dye recoveries from air filters for different groups of dyes.
3.2 Subtask B: Analysis of Dye Drug Room Field Samples
MRI will analyze air filter samples collected during surveys of a number
of dye drug rooms. These analyses will be carried out by the methodology
developed in Subtask A. Both area and personal air sampling will take
place. The monitoring period will consist of one complete work shift at
each drug room site. Two personal air samplers will be worn by one drug
room worker per plant. Field air filter blanks and filter lot blanks
will be collected for background correction and determination of dye
recoveries for certain groups of dyes. Samples of the bulk dyes han-
dled in the drug room will be taken at the end of the shift after pump
shut-off, or during the shift if it is the judgment of the industrial
hygienist that this will not affect the air samples.
Sample analysis will consist of extracting all dyes from the air filters,
measuring and storing the visible absorbance data points from the extract
solutions, and calculating the estimated quantity of dyes present on each
filter based on a physical constant derived from the individual bulk dyes
which were handled during the monitoring period. The uncertainty in the
value of this constant will be proportional to the precision of the indi-
vidual dye absorptivity values. The precision of the absorptivity values
will be documented by performing them in duplicate for the trial plant
analysis. An average airborne dye concentration will be calculated by
dividing the total dye estimate by the volume of air sampled during the
monitoring period.
A-47
-------
Section No.: 4.0
Revision No.: 2
Date: April 30, 1987
Page 1 of 4
SECTION 4.0
PROJECT ORGANIZATION AND MANAGEMENT -
The work assignment organizational chart is shown in Figure 4-1. All MRI
personnel may be reached by telephone at (816) 753-7600.
4.1 Program Management
Mr. Paul Constant, Program Manager, will represent management. He will
be assisted in this effort by Mr. John Hosenfeld, Deputy Program Manager.
Together they will:
• Assure that all necessary resources are available.
- Assure that the Quality Assurance Coordinator (QAC) is fully
informed and involved in the project.
• Assure that all personnel are informed of project QA policy.
• Review all communication from the QAC or QAM regarding the project.
• Assure that any problems, deviations, etc., reported by the QAC or
QAM receive immediate corrective action.
• Assure that the financial standing of the project is fully reported
to the EPA project officer and work assignment manager.
• Review all technical work and reports for overall technical accuracy.
4.2 Quality Assurance Manager (QAM)
Ms. Carol Green, Quality Assurance Manager (QAM), will represent MRI.
• Review the project QA plan to assure that it is consistent with
corporate and client policies and procedures.
•
• Assure MRI management that the facilities, equipment, personnel,
procedures, and records are consistent with corporate and client QA
objectives and requirements by conducting or directing independent
inspections and/or audits.
• Monitor the work assignment QA activities.
• Report unresolved corrective actions to corporate management for
resolution.
A-48
-------
Work Assignment Organization Chart
Section No.: 4.0
Revision No.: 2
Date: April 30, 1987
Page 2 of 4
QA MANAGER
Carol Green
1
QAC OTS "
PROGRAM
Jack Ba(singer
i
QCC
Randy Ay I ing
DIRECTOR
KANSAS CITY
OPERATIONS
D. Sunder man
J. Spigarelli
JL
PROGRAM
MANAGER
Paul Constant
WA LEADER
Don Harbin
ASSISTANT
WA LEADER
Julie Long
ANALYTICAL
Project Staff
Figure 4.1
A-49
-------
Section No.: 4.0
Revision No.: 2
Date: April 30, 1987
Page 3 of 4
4.3 The Quality Assurance Coordinator (QAC)
Mr. Jack Balsinger will serve as the Quality Assurance Coordinator and
will represent program management with respect to quality assurance
He will:
• Assure that all corporate and client QA policies and procedures are
available and understood by the project staff by conducting inspec-
tions and audits.
• Help prepare the project QA plan.
• Approve the project QA plan.
• Assure that the facilities, equipment, personnel, nethods, records,
and controls are consistent with project objectives and requirements
by conducting or directing inspections and/or audits. Inspection/
audit results and corrective action requests will be reported to the
program management, MRI Management, and the QAM.
• Reinspect or audit to assure that appropriate corrective actions were
implemented. Report unsolved actions to the program Management and
the QAM for resolution.
• Conduct additional audits as directed by the program manager and/or
QAM.
• Review and audit data reports and supporting evidence prior to sub-
mission to EPA.
• Prepare QA reports to be submitted to EPA.
4.4 Quality Control Coordinator (QCC)
Mr. Randy Ayling will serve as the QCC. He will:
• Conduct systems audit(s), which include reviewing notebooks,
chromatograms, printouts, and other hardcopy information and report
•the findings to the QAC.
• Prepare performance audit samples.
• Report audit findings to program manager after QAC review and
approval.
• Conduct additional audits as directed by the QAC.
A-50
-------
Section No.: 4.0
Revision No.: 2
Date: April 30, 1987
Page 4 of 4
4-5 Work Assignment Leader/Assistant Work Assignment Leader
Mr Don Harbin will be the work assignment leader. He will be assisted
Dy Ms. Julie Long, assistant work assignment leader. Together they will;
• Help prepare and update the project QA plan.
• Be responsible for training staff where required.
• Be responsible for sample receipt and traceability.
• Enforce instrument calibration and maintenance procedures.
• Maintain document control of lab data, notebooks, records, and other
hard copy information.
• Review and approve all data prior to submittal to EPA.
• Review/validate raw data (e.g., notebooks, forms, strip charts, etc.).
• Ensure that any deviations from protocol are approved, documented
and reported. '
• Be responsible for analytical data traceability.
" !ake "rreci!)(e a?1on on any Prob1ems a™* communicate them in writ-
ing to the QAC and the program and department managements.
• Prepare and submit monthly reports.
• Prepare and submit other reports as requested by the EPA work
assignment manager in conjunction with project staff.
A-51
-------
Section No.: 5.0
Revision No.: 2
Date: April 30, 1987
Page 1 of 1
SECTION 5.0
PERSONNEL QUALIFICATION
Mr. Paul C. Constant and Mr. John Hosenfeld will serve as program manager and
deputy manager, respectively. Mr. Hosenfeld will assist Mr. Constant.
Mr. Constant has also served as program liaison officer and as deputy program
manager on the previous contract. Their credentials were previously submitted
in the proposal for this contract.
Mr. Don Harbin will serve as the Work Assignment Leader. He has significant
experience in the high pressure liquid chromatographic analysis of dyes as
well as trace quantisation methods for the determination of organic compounds
His credentials were previously submitted in the proposal for this contract.
Ms. Julie Long will serve as the Assistant Work Assignment Leader (effective
September 1986). She previously served as Quality Control Coordinator for
this program. She has contributed to a number of research programs requiring
her instrumental skills for the analysis of toxic compounds.
Ms. Carol Green will be the Quality Assurance Manager. She has served in
this capacity since May 1983. Her credentials were previously submitted in
the proposal for this contract.
Mr. Jack Bal singer will be the Quality Assurance Coordinator. His credentials
were previously submitted in the proposal for this contract.
c A¥lin9 w111 serve as Quality Control Coordinator (effective September
1986). He is skilled in spectrophotometric analysis and in conducting per-
formance audits and systems audits.
Dr. Jairus D. Flora, Jr., will perform the statistical analyses of the data
from each plant site. His credentials were previously submitted in the proposal
for this contract.
Mr. Roger Rembecki will serve as Senior Technician conducting the laboratory
analyses for this project. He joined MRI in November 1986. He has been trained
in the dy§ analysis protocols by Julie Long.
A-52
-------
ection NO.: 5.0
Revision No.: 3
Date: April 30, 1987
Page 1 of 3
SECTION 6.0
FACILITIES. EQUIPMENT. CONSUMABLES. AND SERVICES
6.1 Facilities
Sample preparation will be performed in a laboratory designated, in part,
for this project (MRI Lab 324-W). This laboratory is equipped with fume
hoods and an analytical balance contained in a vented glove box. The
windows are covered with low actinic film. The fluorescent lights have
been replaced with red lighting.
Sample analyses will be performed on a Cary 219 spectrophotometer (located
in MRI Lab 324-W) or other equivalent spectrophotometers.
Data file processing will be performed on a Hewlett-Packard 9826 micro-
computer located in MRI Lab 119N.
6.2 Equipment
The equipment used on this task includes:
• Cary 219 spectrophotometer or equivalent; modified to allow an analog
detector signal to be output to an integration device.
• Nelson Analytical A/0 Interface box and related chromatography soft-
ware package (Model 4400).
• Hewlett-Packard Model 9826 microcomputer and peripherals used to run
the software.
• Mettler H20 analytical balance or equivalent; capable of weighing to
the nearest 0.01 mg.
• OuPont P4000 personal monitoring pumps or equivalents.
• Volumetric glassware, Low Actinic.
6.2,1 Calibration
6.2.1.1 The spectrophotometer Is checked on a weekly basis by
qualified MRI personnel. A hoimiurn oxide film traceable
to the National Bureau of Standards is used to ensure
that wavelength readings meet the Manufacturer's speci-
fications. Oxford Spectrochek® QA buffer solutions are
used to ensure that absorbance readings Beet specifica-
tions. If the instrument performance falls outside of
the acceptable range, it will be reported to the Instru-
ment Services Group of MRI and corrective action will
be taken.
A-53
-------
Section No.: 6.0
Revision No.: 3
Date: April 30, 1987
Page 2 of 3
6.2.1.2 The Nelson Analytical chromatography software and Hewlett-
Packard hardware have built-in system checks to monitor
their performance. Error messages wil be displayed if
problems occur. A copy of the specific version of the
software program used for processing the data points
will be archived.
6.2.1.3 The analytical balance is checked before use with weights
that are traceable to or checked against National Bureau
of Standards weights to confirm performance according to
manufacturer's specifications.
6.2.1.4 Personal air sampling pumps will be of the feedback flow-
adjusting type. Each sampling pump will be calibrated
for an airflow of 2 to 2.5 L/min prior to use and checked
afterwards.
6.2.2 Maintenance
Maintenance of the analytical equipment used in this task will be
done according to manufacturer's specifications and at their rec-
ommended frequency. This is summarized in Table 6.1
Table 6.1. Maintenance
Equipment Service Frequency
Spectrophotometer General As needed
Hewlett-Packard 9826 Limited requirements As needed
Balance Cleaning and adjustment for
calibration 1 year
DuPont P4000 personal Replace belts and inlet As needed
monitor pumps filters
6.3 Consumables
All dimethyl sulfoxide (DMSO) used will be A.C.S. reagent grade or better.
All pH 7.0 and pH 3.0 buffer will be reagent grade. All filter spiking
experiments will use filters identical to those used during the field
sampling. Bulk sample containers will be amber pill vials.
A-54
-------
Section No.: 6.0
Revision No.: 3
Date: April 30, 1987
Page 3 of 3
6.4 Services
Health and Hygiene, Inc., will conduct the air sampling at the selected
drug room sites and will perform the following services:
• Preweigh and postweigh air sampling filters, including field air
filter blanks.
• Provide calibrated air sampling pumps for use at the site.
• Take bulk samples of all powder dyes handled in the drug room during
the monitoring period.
• Take bulk samples of all non-dye compounds handled in the drug room
during the monitoring period which might interfere with the analyt-
ical analysis method.
• Provide all survey information regarding site conditions and monitor-
ing period activity by the drug room worker.
• Transfer field air filter samples to new cassettes (not required for
field air filter blanks) after postweighing and ship to MRI along
with original cassettes and support pads.
• Provide suitable documentation for calibration and maintenance of all
air sampling and weighing equipment for inclusion in the work assign-
ment archives.
A-55
-------
Revision No.: 3
Date: April 30, 1987
Page 1 of 8
SECTION 7.0
DATA GENERATION
7.1 Experimental Design
The sampling design will be prepared by EPA. A flow chart of the steps
involved 1n the analytical nethod is shown in Figure 7-1.
7.2 Sample Collection (Health and Hygiene, Inc.)
7.2.1 Area Sampling
Area sampling will be performed using samplers operating under
critical flow conditions of 5 to 8 L/min. One area sampler will
be placed near the drug room weighing station and another will be
located at a bulk dye storage area. The sampling period will be
for the entire 8-h shift.
7.2.2 Personal Sampling
One drug room worker will be monitored at each plant. The worker
will wear two personal air samplers. The 37-mm open-face sampling
cassettes will be operated at air flow rates of 2 to 2.5 L/min.
When sampling, the inlet of the cassettes will be pointed downward
so that only airborne material will be collected during the 8-h
shift.
7.2.3 Blanks
Filter lot blanks and field air filter blanks will be provided by
Health and Hygiene. Filter lot blanks will be filters from the
same lot as those used to collect field samples, but which will
not be sent out to the drug room site. Field air filter blanks
will be filters which are handled in the same manner as the sampl-
ing filters except that no air will be drawn through them.
7.2.4 Bulk Dye Samples
. Small samples of each powder dye handled in the drug room during
the monitoring period will be taken and labeled with unique bar-
code stickers provided by MRI. An identical bar-code sticker will
be placed on the bulk dye inventory sheet along with the full dye
name, manufacturer, and lot number. Bulk samples will be taken
with disposable spoons (one for each dye).
A-56
-------
Revision No.: 3
Date: April 30, 1987
Page 2 of 8
Field Air Filter Samples
and Air Sampling Cassettes
Spike blank air
filters with a known
dye mixture (e.g.,
spiking standard
solution) at levels
^bracketing the amount
estimated on the
personal air filters.
Pull air through the
filters for 6 to 8 hr.
Extract dyes from each filter
or cassette using a known
volume of dye solvent.
.Prepare solutions of
each individual dye
at known concentrations
in dye solvent.
\
Prepare reference
standards using
known amounts of the
same spiking standard
solution used to
spike the blank air
filters.
dye recovery -
T cal"culTtioin
Compare total
absorbance of each
spiked filter
extract to that of
the corresponding
reference standard.
Calculate the average
percent recovery, R.
Scan solution bverw)0-330
range on spectrophotometer
store spectrum.
Integrate area beneath the
spectrum to obtain total
.absorbance
nm
and
individual
air filter or
cassette extract
dyes
Calculate the total
dye estimate on the air
filters and sampling
cassettes.
\
Calculate the spectral
absorptivity constant
(ae) for each dye.
i
Calculate the weighted
average spectral ab-
sorptivity constant
(a.) for the dye analysis
set.
Correct air filter dye
estimates by using dye
recovery and dye purity
information. Correct
cassette dye estimates
by using only dye purity
information.!
Obtain estimate of mg
total dye/m3 of air.
Figure 7-1. Flow chart for the estimation of total dyes on an air filter.
A-57
-------
Revision No.: 3
Date: April 30, 1987
Page 3 of 8
7.2.5 Collection Conditions
Temperature, relative humidity, and barometer measurements will
be monitored and recorded during the test. A general description
of the airflow characteristics in the drug room will be recorded.
The sampling time interval will also be documented.
7.2.6 Shipping and Handling
After gravimetric analyses of the field air filters have been
performed at Health and Hygiene, Inc., the air filters, original
cassettes and support pads, bulk dye samples, field air filter
blanks, filter lot blanks, and the field data forms will be
shipped to MRI by overnight courier. Bulk dye samples will be
shipped separately to minimize the chances of contamination.
7.3 Sample Traceability
Tracking of field air filter samples, field air filter blanks, and bulk
dye samples will be achieved using the field data forms (Figure 7-2).
During each phase of field testing (e.g., air sampling, gravimetric analy-
sis) all samples will be assigned/identified using a unique sample identi-
fication number.
7.4 Laboratory Analysis Procedures
See Appendix B for the analytical protocol.
7.5 Internal Quality Control Checks for Sample Analyses
7.5.1 General
New and current lots of reagents are checked prior to use.
7.5.2 Calibration
Proper Instrument performance will be confirmed and documented
(see Section 6.2.1).
A-58
-------
Revision No.: 3
Date: April 30, 1987
Page 4 of 8
PERSONAL AREA AIR SAMPLING DATA SHEET
Plant ID No: Saap.
Flotoeter Model Nuafeer:
Date:
Tiae:
Tiae:
Sample ID Nutter:
Sample Duration: (Bin)
Puop Flow Rate: (L/un)
Sample Air Voluae: (B3)
issssssssssssssszsssssssss
Signature:
Calculations Checked by:
Sent by: Date:
Sent by: Date:
ling Performed by:
I Job Title/Work
SAMPLING EQUIPMENT AND CAL
Floweter Sen
te: Calibration Tr
i
FIELD SAMPLING DATA
left [] right []
i
TRACEABILITY RECORD
Tiae: > Rec'd
Tiae: — > Rec'd
JDate CHo/Da/YD
Duties: _
JBRATION
al Nuaber:
aoeable to:
left [] right []
[Date: j
JDate:
gggM—— »»-gl-— •— - — -———--
by: Date: Tiae: :
by: Date: Tine:
Figure 7-2
A-59
-------
Revision No.: 3
Date: April 30, 1987
Page 5 of 8
Plant ID No:
STATIONARY AREA AIR SAMPLING DATA SHEET
=s=s::ssss:ss=s
Sampling Performed by:
jDate (Mo/Da/Yr)
SAMPLING EQUIPMENT AND CALIBRATION
FloMoeter Model Number:
Flc
eter Serial NuBber:
Fl
ter Calibration Date:
Calibration Traceable to:
Sampling Punp Model No:
Sampling Piop Serial No:
Pre-aaopling Flowrate:
Date:
TIM:
Flowrate:
Date:
Sicnaturefa):
Sample ID Number:
FIELD SAMPLING DATA
ISSSSSSSSSSSSSSSSSSSS
Sampler Location:
Sample Start Tine:
Sample Stop Time:
Sample Duration: (aun)
Flap Flow Rite: IL/ain)
Sample Air Volua*: (m3)
•~**""*-'--r~r"r""~~"~~~~~*
Signature:
Date:
CalAulations Checked by:
BSSSSSSSSSSSSSS55SSSSSSSSS
i°
BSSSS:
Date:
iSSSSSSSSSSSMSSSSSS:
TRACXABILTTY RECORD
BSSSSSSSSSSSSSSSSSSfSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
Sent by:
Date:
Tiae:
Rec'd by:
Date:
SSSSSSS!
Tiae:
Sent by:
Date:
Tint:
Rec'd by:
Date:
Ti*e:
Figure 7-2 (continued)
A-60
-------
Revision No.: 3
Date: April 30, 1987
Page 6 of 8
ANALYTICAL DATA
p»===«=s=s=~::==ss=ss:=:zssz===sssssrssssss==s::
1 GRAVIMETRIC ANALYSIS (HeaJ
1
Sample ID Number:
> Filter Preweicht:
Filter FOBtweight:
i
Sample Height:
1
Blank Correction:
Adjusted Weight:
i
Personal Filters
1th and Hygiene)
:ss=s:=s=====s:s==s::s:
Area Filters
Blank Filters
L 1
!
, Signature!*): Date: Calculations Checked by: \
VISIBLE SPECTROPHOTCKETRIC ANALYSIS (NU)
Filter Extract. Date:
i
Data File Number: •
Sample Prepared by: ;|
Total Absorbance:
1
jCorr. Total Absorbance:
Absorptivity: (As)
Dye Estimate: (ug)
i
: !
i j
i : :
i
, Avg. Recovery: (ft) |i
i '•
Corr. Dye Est: (ug)
;Est. Airborne Dyes: (usVm3)i-
. i
Data Reference Number: ;
i
!
!
i
i
!
;
:
!
1
1
I
Signature (*): .Date: Calculations Checked by: !
ssssss*ssssssssss=:rsssss±sssssssss
ssssssssssssssszssssssssssssssssssssss:
Figure 7-2 (concluded)
A-61
-------
Section No.: 7.0
Revision No.: 3
Date: April 30, 1987
Page 7 of 8
7.5.3 Internal QC Samples
With each batch of samples, appropriate QC samples will be in-
cluded so the quality of the sample data can be assessed. These
QC samples include reagent blanks, field air filter blanks, and
spiked filter controls for determining extraction efficiency.
7.5.3.1 Reagent blanks: At least one reagent blank will be
analyzed each analysis day to check for solvent inter-
ferences. No filters will be used for this determina-
tion.
7.5.3.2 Field air filter blanks: The field air filter blanks
are filters from the same lot as the air sampling fil-
ters. Field air filter blanks will be subjected to the
same handling procedures as the field air filter samples
except that no exposure to the drug room environment will
be allowed. At least one field air filter blank will be
collected and analyzed for each drug room site monitored.
7.5.3.3 Spiked filter controls: Filter lot blanks will be spiked
with known amounts of a dye mixture comprising a subset
of the total number of dyes handled in the drug room.
This dye mixture (i.e., the spiking standard solution)
will be composed of the individual dyes which account
for at least 80% of the total quantity of dye handled by
the drug room worker during the monitoring period. The
relative amounts of the component dyes in the spiking
standard solution will reflect their actual usage during
the monitoring period. Spiked filter controls will be
prepared in replicate to check the precision of the re-
covery experiments.
7.6 Systems and Performance Audits
7.6.1 Systems audits: Systems audits by the QAC or QCC shall include:
• Inspecting facilities and equipment for adequacy, appropriate-
ness, and safety during use.
• Reviewing actual practices versus written procedures and pro-
tocols.
• Inspecting the records of maintenance and calibration.
• Inspecting QC practices.
• Preparing and submitting a report with recommended corrective
actions to the QAC, and after approval, to the work assignment
leader, program management, and the QAM.
A-62
-------
Section No.: 7.0
Revision No.: 3
Date: April 30, 1987
Page 8 of 8
7.6.2 Performance Audits
A performance audit sample (PAS) is designed to check the opera-
tion of the equipment as well as the analytical method and data
reduction procedures. Performance audit samples will be prepared
]nf?^enily ty-K!e Q/Vr QCC USin9 the bu'k d*es and win'be
ana yzed along with regular samples. The audit samples will be
included periodically (beginning, middle, and end) during analysis
Su4nn h ^H™™ '?* ^g^ durin9 f™™ «1r f11ter analysis,
during bulk dye analysis, and during the dye recovery experiments)
I^%nlIyS%riVArep°rl!:0 the QCC or <>AC the total absofbance of '
fir P^'DA! ?AC ?K QCC Wil1 calculat* a "found" concentration
for each PAS using the spectral absorptivity constant of the
?a« 15 ?hf(fr; JVff I°Und concentration does not agree within
fr?SnJ i? J ! JU? bu ?,dy.e """^ration, one or more of the
actions listed below will be taken:
l' Ic?n2AfK0r QCC ?ll !up?^ another audU samPle f°r analysis,
using the same bulk dye(s) as before.
2. Calibration check of the Gary 219 spectrophotometer to verify
specifiwtiJns ^ wavelength r«qui>ements fall within
3. Confirm the spectral absorptivity constant of the dye(s) used
for the PAS by repeat analysis.
JILF-ftJT6" aud: VfP16 results and any corrective actions
taken will be reported to the work assignment leader, program
management, and the QAM. ' program
oAhpw ais° b «
QAC believes the analysis procedure has changed, (2) analytical
problems are suspected, (3) the MRI work assignment leader or the
EPA work assignment manager requests samples.
7.6.3 QAC Data Audits
3111 be ?ondu^ed or directed by the QAC by reviewing,
rr ,rr nJppPOVJ9B *" r?ports and supporting evidence for
accuracy and QA compliance prior to report submittal to EPA
A-63
-------
ec ion o.: g.o
Revision No.: 3
Date: April 30, 1987
Page 1 of 2
SECTION 8.0
DATA PROCESSING
8.1 Collection
itlOn Will Ut'M'i'Ta hn+h -- 1 •
acquisition
8.2 Data Reduction
8-3 Data Validation
The data validation process will include:
"d COBputer "™9r»S «d decanting the
Validating and checking electronic data transfer
Screemng data for consistency by a second project staff member.
Checking calculations, randomly.
Performing outlier checks.
-
A-64
-------
Section No.: 8.0
Revision No.: 3
Date: April 30, 1987
Page 2 of 2
Examining QC data and QC checks.
Maintaining records of reviews and validation.
comPleteness' representativeness, and
• Reviewing and approving of all records by the work assignment leader.
• Reporting protocol deviations and assumptions with the results.
The work assignment leader will be responsible for assuring data validity.
8.4 Storage
Raw data will be documented in laboratory notebooks, on data forms or
printer paper, as strip chart recordings, or as hardcopy originals from
S?iehiC.!2!?S 2r.dlsk!' Permanent storage of work assignment records
will be archived in a formal project file (SOP-QA7)
A-65
-------
Section No.: 9.0
Revision No.: 2
Date: April 30, 1987
Page 1 of 3
SECTION 9.0
DATA QUALITY ASSESSMENT
This analytical method gives an estimate of total dye content on an air filter
and should not be considered as a quantitative determination of the amount of
dyes present. The precision of the analytical method should meet expected
standards for a spectrophotometric method which involves the extraction of
rmcrogram quantities of analytes from an air filter. The objectives of pre-
cision for this method will be to obtain dye recoveries for replicate spiked
filter samples which have relative differences ± 25% of each other. The ob-
jective for accuracy will be to obtain total dye estimates on replicate spiked
filter samples which have relative errors ± 50% of the actual dye present on
the air filter. Average extraction efficiencies for the dyes using spiked
filters should fall within the range of 60 to 140% recovery to yield meaning-
TU I Q3 u3 .
9.1 Precision
The precision of the analytical method will be determined by analyzing
replicate spiked filter samples and calculating their respective percent
recoveries. Percent relative difference, R.D., will be calculated as
follows:
" x 100
where R! = the % recovery for one replicate
Ra = the % recovery for the other replicate
R = the average % recovery calculated from Ra and R2
9.2 Accuracy
The accuracy of the analytical method can only be established for known
dye spike samples. Accuracy may be indicated by comparing the total dye
estimate for spiked filter controls to the actual amount spiked on the
filters. Accuracy will be measured by calculating the relative error
R.E., of the total dye estimate: '
R.E. (%) = " °ACT x 100
ACT
A-66
-------
evision NO.: 2
Date: April 30, 1987
Page 2 of 3
where DE$T = the total dye estimate in ug
DACT = the actual total dye quantity spiked in pg
9.3 Uncertainty
The uncertainty of the total dye estimation will be primarily dependent
on the specific group of dyes being analyzed. The dye estimation is
based on the weighted average absorptivity of all of the dyes handled
during the monitoring period. In general, the uncertainty of the dye
estimate will be proportional to the standard deviation of the individ-
ual dye absorptivities. For this reason it is not possible to establish
a specific uncertainty value for the total dye estimate.
Probable errors In the total dye estimate can be obtained by using a
statistical computer program that selects dye mixtures from the given
group of dyes. Specific dye handling Information is input into the
program to weight the dye selection process. As a result, the more
heavily used dyes have a higher probability of being selected than the
minor use dyes. In this manner the errors associated with various sub-
sets of the entire group of dyes can be approximated. Additional data,
such as the number of dye weighings or dustiness observations, can then
be used to focus in on dye subsets which are more likely to occur on the
air filter.
9.4 Recovery
The efficiency of the filter extraction procedure will be indicated from
the recovery results of the spiked filter analyses. Dye recovery will
be determined by a direct comparison of the spiking solution (e.g.,
reference standard) to the solution obtained from extracting the spiked
filter. r
R (%) = /il x 100
*STD
where Afil = total absorbance of spiked filter extract
total absorbance of reference standard
9.5 Traceability of Instrumentation
All collection and Measuring instrumentation will have a unique Identi-
fication number. Maintenance, calibration, and use logs will be main-
tained.
A-67
-------
Date: April 30, 1987
Page 3 of 3
9-6 Traceabnity of Samples
^linn^'f^I1 hav? a Uni'que Verification number along with infor-
mation about the worker being monitored, the plant site monitor nn i«
cation, exposure time and conditions, collection deJice,' J^0™9 1o
9-7 Traceability of Data
9.8 Completeness
A-68
-------
•
Revision No.: 2
Date: April 30, 1986
Page 1 of 2
SECTION 10.0
CORRECTIVE ACTION
Some^of the types of problems and corrective actions to be taken are listed
8-l Performance/Systems Audits
If problems are detected during an audit:
• The auditor shall notify the person responsible, the work
leader, and the QAC of the problems) aSd any ttiS(O
10.2 Loss of Data
rented frSlrn™-"?1**" *2 KT' the P~'le»/action taken is docu-
TC . 2 «>e project records; the work assignment leader then pre-
" " 'n"m/l>c«°" «"»n Kmo to the QAC and the Irog
anage
- *
10-3 Significant QA Problems
In general, the work assignment leader shall identify technical problems.
A-69
-------
Section No.: 10.0
Revision No.: 2
Date: April 30, 1986
Page 2 of 2
The work assignment leader prepares and sends a problem memo to the
QAC and program manager; if the problems are significant, the action
is determined collectively. __
The action taken is documented in the project records.
The problem and action taken is reported to the EPA work assignment
manager.
A-70
-------
Date: April 30, 1987
Page 1 of 2
SECTION 11.0
DOCUMENTATION AND REPORTING
11.1 Documentation
• All documentation shall be in permanent ink or on computer printouts.
• Corrections will be performed as follows: Draw a single line through
an incorrect entry so that the original entry remains legible. Add
the correct entry; then explain, initial, and date the correction.
• New information may be added to original raw data. It will be
initialed, dated, and explained.
• All deviations from standard operating procedures (SOPs), procedures,
and protocols will be documented.
• All assumptions and interpretations will be documented.
• Strip charts, magnetic tapes, etc., will be labeled with a format
identifier, the date, the ID(s) of the sampling equipment, and the
name of the person responsible for the data recording equipment.
Hardcopies of all nagnetic data will be generated for archiving
purposes.
11.2 Document Control
• Raw data will be documented in laboratory notebooks, on sampling
forms, on analytical forms, on printer paper, as hardcopies from
magnetic tape, and as strip chart recordings.
: A raw data packet for each drug room site monitored will be generated,
along with data tracking forms to document the existence and flow of
data through the data processing cycle.
• All project-related documents will be maintained by assigned project
staff until archived.
11.3 QA Reports to Program Management
The QAC, in cooperation with the work assignment leader, shall identify •
critical areas of the project which will be subject to inspection. The
inspection will include a review of:
• Staff credentials.
• Equipment maintenance and calibration records.
• Equipment performance.
A-71
-------
Revision No.: 2
Date: April 30, 1987
Page 2 of 2
• Documentation practices.
• Recordkeeping practices.
• Adherence to protocols, SOPs, and the QA project plan.
• Assessment of data accuracy, precision, and completeness.
The results of inspections and audits will be reported to the work
assignment leader, the program manager, and QAM.
11.4 Report Design
Progress, draft final, final reports, and QA summary reports will be sub-
mitted in accordance with the provisions for reporting in the contract
Verbal status reports will be made biweekly to the work assignment leader
A-72
-------
o.: ppen ix A
Revision No.: 2
Date: April 30, 1987
Page A-l of 4
APPENDIX A
STANDARD OPERATING PROCEDURE FOR CHECKING THE CALIBRATION
OP THE GARY 219 SPECTROPHOTOMETER
A-73
-------
Section No.: Appendix A
Revision No.: 2
Date: April 30, 1987
Page A-2 of 4
STANDARD OPERATING PROCEDURE FOR CHECKING THE CALIBRATION
OF THE CARY 219 SPECTROPHOTOMETER
Absorbance and wavelength verification of the Cary 219 spectropho-
tometer (or equivalent) is performed and documented weekly by URL The ab-
sorbance is monitored using Oxford Spectrochek® QA solutions (or equivalent).
The wavelength is monitored using a NBS traceable holmium oxide film. If the
instrument does not pass verification, the Instrument Services Group of MRI
will be notified and corrective action will be taken. Verification .documents
will remain on file at MRI.
The following steps outline the operating procedure for conducting
the spectrophotometric calibration check:
I. Wavelength Verification via Holmium Oxide Film
A. Turn main power switch to "on" position. Allow Vis-UV light source
to warm up for at least 20 min.
B. Initially, an air versus air scan will be performed to zero the
instrument. The conditions for setup are identical to those listed
in Appendix B, Section 13.1.2 except for the following changes:
- Slit; 1.0 nm
- Scanning rate: 0.5 nm/s
• Chart display: 5
C. Close the covers to the cuvette compartments on the spectrophotom-
eter.
0. Set the upper wavelength limit at 750 nm using the wavelength 1 dial
and set the lower wavelength limit at 250 nm using the wavelength 2
dial.
E. Using the "scan" dial, turn dial (+) or (-) to set the wavelength
at 750 nm.
F. . Turn the timer node knob fully clockwise, then back to "wavelength."
G. Turn the autobaseline knob fully clockwise to the "record" position
and hold for a couple of seconds, Baking sure the red recording
light comes on.
H. Adjust the balance knob to give an absorbance reading of 0.1000 (or
as close as possible) on the digital display.
I. Insure measurement dial 1s set on serial.
A-74
-------
Section No.: Appendix A
Revision No.: 2
Date: April 30, 1987
Page A-3 of 4
J. Press "step" button and scan the wavelength range.
K. When the scan is finished, the instrument will return the wave-
length to 750 nm.
L. Turn the pen off.
M. Turn the timer mode knob fully counter-clockwise to "off" position.
Line up pen on a dark graduation.
N. Adjust the balance knob so the digital readout shows 0.000 (or as
close as possible).
0. Turn timer mode knob fully clockwise, then back to "wavelength."
P. Turn measurement knob to overlay.
Q. Turn pen on and press "step" button to begin scan.
R. When scan is finished, turn the pen off. Watch to be sure the
paper retracts properly.
S. Turn the timer mode knob to "off."
T. Remove the sample cover, place the holmium oxide film into the sam-
ple turret, and close the lid.
•
U. Check to be sure the pen has realigned itself to the same starting
point as in step M.
V. Turn timer mode knob fully clockwise, then back to "wavelength."
W. Turn pen on and press "step" button to begin scan.
X. When the scan has finished, turn the pen and the timer mode knob to
off positions.
Y. Remove the holmium oxide film.
•
H- Absorbance Verification via Oxford Spectrochek® QA Buffer Solutions
A. Initally, a water versus water scan will be performed to zero the
instrument, using two 1-cm path width cuvettes filled with deionized
water. Clean the outer surface of each cuvette with an appropriate
tisue to remove any smudges.
A-75
-------
Section No.: Appendix A
Revision No.: 2
Date: April 30, 1987
Page A-4 of 4
B. Place one cuvette in the reference compartment and place the other
cuvette in the sample compartment.
C. Dial in 690 nm for the upper wavelength limit and 300 nm for the
lower wavelenght limit using the wavelength 1 and 2 dials, respec-
tively. Using the "scan" dial, turn the dial to (-) to set the
wavelength at 690 nm.
D. At this point, follow the steps outlined in F through S for the wave-
length verification via the holmium oxide film. (Switch scan rate
to 2.0 nm/s at this point.)
E. Remove the cuvette from the sample compartment. Fill with solution
no. 2, rinsing the cuvette prior to filling with * 2 ml of the same
solution. Wipe the cuvette surface and replace in sample compartment.
F. Reset the wavelength range to scan from 690 to 400 nm.
G. Fully turn the timer mode knob clockwise, then back to "wavelength."
H. Turn the pen on and press "step" button to start the scan.
I. When the scan is finished and the wavelength has returned to 690 nm,
turn the pen and the timer mode knob off.
J. Open the sample compartment and remove the sample cuvette. Rinse
the cuvette with solution no. 1 and fill. Replace into sample com-
partment and close lid.
K. Repeat steps G through I.
IT. Remove the sample cuvette, rinse with solution no. 4 and fill.
Place back into the sample compartment and close lid.
M. Reset the wavelength range to scan from 400 nm to 300 nm. Using
the "scan" dial, turn the dial to (-) to set the wavelength at
400 nm.
N. Repeat steps G through I.
•
0. Remove the sample cuvette and repeat steps G through I using
solution no. 3.
A-76
-------
ec ion o.: ppen ix
Revision No.: 3
Date: April 30, 1987
Page B-l of 20
APPENDIX B
ANALYTICAL PROTOCOL
A-77
-------
ec ion o.: puen ix
Revision No.: 3
Date: April 30, 1987
Page B-2 of 20
ANALYTICAL METHOD FOR THE ESTIMATION OF TOTAL DYES IN
TRACE QUANTITIES ON AIR SAMPLING FILTERS
1.0 SCOPE AND APPLICATION
1.1 This method has been developed for the estimation of trace quan-
tities of groups of dyes from the following dye classes: acid,
basic, direct, disperse, and reactive. Dyes from other major
classes have not been investigated.
1.2 This method yields an estimate of the total quantity of a group
of dyes present on an air sampling filter. Information about the
quantity of each individual dye on the air filter is not possible
with this method.
1.3 This method is suitable for use by technicians possessing nominal
experience with spectrophotometric equipment and procedures.
2.0 SUMMARY OF METHOD
This method describes the procedures for estimating the total quantity
of a group of dyes present on an air sampling filter. A general dia-
gram of the method is shown in Figure B-l.
The analysis procedure consists of extracting the dyes from each air
filter using a minimum volume of buffered dye solvent. An aliquot of
each filter extract is passed through a 0.45 urn Gelman Acrodisc (or
equivalent) and transferred to the measuring cell in the spectrophotometer.
The visible absorption spectrum of each extract is obtained and digitized
by means of an A/D converter box interfaced with the spectrophotometer.
Air sampling cassettes and one or more field air filter blanks are also
extracted and scanned. All spectra are stored on floppy disks for fu-
ture data manipulation.
Spectra of each individual dye handled in the drug room during the air
monitoring period are obtained by analyzing known solutions of bulk dye
samples collected at the plant site. These dye spectra are then used
to form a data base for estimating the total quantity of dye on the air
filters (and sampling cassettes) taken during that monitoring period.
Spectra obtained on identical bulk dyes at different plants will not be
employed so as to avoid possible lot-to-lot variations in absorption
characteristics.
A-78
-------
ec ion D.: ppencnx
Revision No.: 3
Date: April 30, 1987
Page B-3 of 20
Field Air Filter Samples
and Air Sampling Cassettes
Spike blank air
filters with a known
dye mixture (e.g.,
spiking standard
solution) at levels
^bracketing the amount
estimated on the
personal air filters.
Pull air through the
filters for 6 to 8 hr.
Extract dyes from each filter
or cassette using a known
volume of dye solvent.
Prepare solutions of
each individual dye
at known concentrations
in dye solvent.
i
Prepare reference
standards using
known amounts of the
same spiking standard
solution used to
spike the blank air
filters.
Scan solution over 800-330 nm
range on spectrophotometer and
store spectrum.
dye recovery
r cafculTti'on
Compare total
absorbance of each
spiked filter
extract to that of
the corresponding
reference.standard.
Calculate the average
percent recovery, R.
Integrate area beneath the
spectrum to obtain total
absorbance (A-.,V
individual
air filter or I
cassette extract I
Calculate the total
dye estimate on the air
filters and sampling
cassettes.
Correct air filter dye
estimates by using dye
recovery and dye purity
information. Correct
cassette dye estimates
by using only dye purity
information.I
Obtain estimate of mg
total dye/m3 of air.
dyes
Calculate the spectral
absorptivity constant
(a ) for each dye.
Calculate the weighted
average spectral ab-
sgrptivity constant
(ac) for the dye analysis
set.
Figure B-l. Flow chart for the estimation of total dyes on an air filter.
A-79
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-4 of 20
The total sample absorbance in the visible wavelength region is used to
estimate the total quantity of dye present on each air filter and its
corresponding cassette. This entails electronically integrating the
area beneath the absorption spectra of each filter or cassette extract
and each individual dye. The area values are readily converted to total
absorbance (ATOT) by using an appropriate factor (i.e., area count value
•f 106 area counts/ absorbance unit).
Spectral absorptivity (a ) constants for each individual dye used dur-
ing the monitoring periofl are calculated based on Beer's law. From the
individual a values, and dye handling information, a weighted average
spectral absorptivity constant (a$) for the group of dyes is calculated.
The total amount of dyes contained on each air filter, and its corre-
sponding cassette, is then calculated using this weighted average. The
dye estimates are corrected wherever appropriate for the amount of ac-
tive ingredient in each dye (i.e., purity) and for the average recovery
(R) of dyes off the air filters.
3.0 INTERFERENCES
3.1 Because of possible reactivity with the dye solvent at the estab-
lished pH (pH 7), some basic dyes may not be quantifiable using
the standard analytical procedure. Use of pH 3 dye solvent is
recommended if the bulk dyes handled at the plant site are pre-
dominantly of the basic class.
3.2 Due to the nature of the analytical technique used, this method
is not particularly susceptible to other interferences except
those from other dyes, i.e., cross-contamination. Therefore,
after each use, glassware is immediately rinsed with methanol to
remove all traces of dye and dye solvent. The glassware is then
washed with soap/H20, rinsed with deionized water followed by
acetone, and allowed to air dry prior to re-use.
3.3 Since fluorescent brighteners are not to be included in the total
dye estimate, the lower wavelength limit of the absorbance scan
may have to be shifted upward if brighteners were handled during
the air sampling period. If this change in the wavelength scan
range should be necessary, all the individual dye and filter/
cassette extract solutions in the dye analysis set must be scanned
over this same wavelength range.
4.0 SAFETY
All manipulations made with dyestuff samples are performed in a fume
hood or glove box. Gloves and other appropriate safety apparel are
A-80
-------
Revision No.: 3
Date: April 30, 1987
Page B-5 of 20
worn at all times. Solid and liquid waste is disposed of in the proper
manner.
5.0 APPARATUS AND MATERIALS
5.1 Solution Preparation
5.1.1 500 and 1,000 ml graduated cylinders
5.1.2 1 gal glass bottle with 10 ml Repipettor®
5.1.3 Volumetric pipets (TO) - 2, 3, 5, 6, 8, 10, and 20 ml
5.1.4 Volumetric flasks - 10, 25, 50, 100, 250, 500, and 1,000 ml
(low actinic)
5.1.5 Disposable pipets
5.1.6 Beakers - 100 ml
5.1.7 Filters - 0.45 urn Gelman Acrodisc (or equivalent)
5.1.8 Glass jars (amber) - 4 oz with DMSO-resistant lid liners
(optional)
5.1.9 10-mL disposable syringe (Luer tip)
5.1.10 Stainless steel forceps
5.2 Balance - Analytical; capable of accurately weighing to 0.01 mg
5.3 Shaker - Capable of shaking 4-oz jars at 1 oscillation/s. If a
wrist-type shaker is employed, DMSO-resistant lid liners must be
used on the glass jars.
5.4 Ultrasonic bath
5.5 Spectrophotometer/data storage system
5.5.1 Spectrophotometer - Dual beam instrument capable of scan-
ning in the visible wavelength region (800-330 nm). The
Spectrophotometer must have a 1 or 10 V analog signal out-
put, or allow such a signal to be obtained (such as by
using the strip chart pen signal).
5.5.2 Cuvettes - Standard 1 cm pathlength.
5.5.3 Nelson Analytical Model 4400 Chromatography Data System,
* or equivalent.
5.5.4 Nelson Analytical A/D interface box, or equivalent.
5.5.5 Magnetic media for data storage - 5-1/4 in. floppy disks,
or equivalent.
A-81
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-6 of 20
6.0 REAGENTS
6.1 Dimethyl sulfoxide (DMSO), ACS certified grade
6.2 Buffer, pH 7 and pH 3 - reagent grade (uncolored)
6.3 Dye solvent (pH 7) - prepared by adding 9 parts DMSO to 1 part
pH 7.0 buffer
6.4 pH 3 dye solvent - prepared by adding 9 parts DMSO to 1 part pH
3.0 buffer
7.0 METHOD VALIDATION (Average Recovery Determination, R)
The analytical method is validated using a subset of the dyes handled
in the drug room during the monitoring period. This subset is composed
of the individual dyes used most during the monitoring period (based on
weight). The component dyes in the subset should represent at least
80% of the total quantity of dye handled and be in proportion to their
usage.
Duplicate blank filters are spiked at two levels with a solution of the
dye subset (i.e., spiking standard solution) so as to bracket the absorb-
ance of the personal air filters. After drying, clean, humid air is
pulled through the filters using similar conditions (time, flow rate)
used in the field. Validation is confirmed if the average recovery for
the spiked filters falls in the 60 to 140% range.
8.0 SAMPLE RECEIPT AND STORAGE
8.1 Sample Receipt
8.1.1 Samples received from the field consist of:
1. Personal and area air sampling filters, which were
removed from their original cassettes and stored in
fresh cassettes after gravimetric weight determina-
tion.
2. The original air sampling cassettes used during the
monitoring period.
3. Field air filter blanks in cassettes, which were
carried in the field along with the actual air fil-
ter samples.
A-82
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-7 of 20
4. Filter lot blanks (unexposed air filters for use in
the recovery experiments).
5. Individual bulk dyes.
8.1.2 Upon receipt, each filter/cassette sample is examined and
logged-in using Form A-0. Individual bulk dyes, identified
by bar-code labels, are logged-in via use of an appropriate
bar-code reader.
8.2 Sample Storage
8.2.1 Unless instructed otherwise, samples are stored in the dark
in room 324W at room temperature in their original packing
containers until the analysis is completed. After analysis,
bulk dye samples are retained at the discretion of EPA.
8.2.2 Although field air filter samples should be analyzed as
soon as possible after receipt at MRI, analysis will not
be performed without (a) a performance audit sample (PAS)
being analyzed concurrently, and (b) without knowledge of
what type of dye solvent should be used to perform the
analyses (i.e., pH 7 or pH 3). When air filters arrive at
MRI, the EPA work assignment manager will be contacted.
Air filter analysis will be delayed until (a) the indi-
vidual bulk dyes (which are used to prepare the PAS)
arrive at MRI and (b) basic dye information regarding the
dyes handled during the monitoring period is received.
The above measures will be observed unless instructions to
the contrary are given by EPA.
9-0 EXTRACTION OF PERSONAL AND AREA FILTERS. CORRESPONDING AIR SAMPLING
CASSETTES. AND FIELD AIR FILTER BLANKS
9.1 Using stainless steel forceps, carefully transfer the air filter
and filter support from the cassette holder to a 4-oz amber glass
.jar. (Note: Each original air sampling cassette will contain
only a filter support.)
9.2 Using the Repipettor® calibrated to 8.0 mL, pipet approximately
2 mL of the dye solvent into the cassette holder, dispensing the
remaining ^ 6.0 mL of dye solvent into the jar.
9.3 Cap the cassette holder tightly and shake vigorously for 30 s.
Transfer the cassette rinse solution to the 4-oz jar using a dis-
posable glass pipette. Cap the jar securely.
A-83
-------
FORM A-0. SAMPLE RECEIPT FORM
8856-A(01)
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-8 of 20
Plant ID:
Test Date:
Date Received:
Received by:
Were all samples received in good condition? Yes No
Sample Storage:
(circle one)
Sample ID No.
t
San
Original Air
Filter Cassette
pie Description (Check One)
Personal
Air Filter
Area
Air Filter
.
Held Blank
Air Filter
Were any blank filters received? Yes No
Brand and Lot No.:
(circle one)
A-84
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-9 of 20
9.4 Place the jar in a shaker for 30 min. The shaker must oscillate
at least once per second. After 30 min, visually examine the air
filter and/or filter support. If the dye extraction appears in-
complete, place the jar in an ultrasonic bath for a maximum of
10 min.
9.5 The extract solution must be analyzed within 45 min of Initial
contact with the dye solvent (I.e.. Step 9.2).Dye analysis
schedules (Form A-l) are prepared and followed for each set of
filters/cassettes analyzed.
9.6 Only when deemed appropriate after consultation between MRI and
EPA, should the pH 3 dye solvent (Step 6.4) be used to perform
the extractions.
10.0 INDIVIDUAL DYE SOLUTION PREPARATION
Note: Dye solutions must be analyzed within 45 min of initial
contact with the dye solvent (Step 10.2^!Dye analysis schedules
(Form A-l) are prepared and followed for each set of bulk dyes
analyzed.
10.1 Accurately weigh (to the nearest 0.01 mg) approximately 40 mg of
the bulk dye sample into a pre-tared 100-mL volumetric flask (low
actinic). Record the weights on the Individual Dye Solution
Preparation Form (Form A-2).
10.2 Add 50-70 ml of dye solvent to the flask and place in an ultra-
sonic bath for 10 min. After the flask has cooled to room
temperature, dilute to volume with dye solvent and mix well.
10.3 Pipet 10.0 ml from the flask in Step 10.2 to a 250-mL low actinic
volumetric flask. Dilute to volume with dye solvent and mix well.
10.4 Repeat Steps 10.1 through 10.3 for every dye in the dye analysis
set.
10^.5 If pH 3 dye solvent was used in Steps 9.2-9.5, use the pH 3 dye
solvent (Step 6.4) to prepare the dye solutions.
11.0 PREPARATION OF SPIKED FILTER CONTROLS AND BLANKS (Forms A-3 and A-4)
11.1 Based on the average total dye estimate for the personal air fil-
ters (obtained from Step 16.5 and using Section I of Form A-3),
calculate the total dye level which is 130% of that value. This
calculated dye level is the target total dye concentration for
the spiking standard solution.
A-85
-------
FORM A-l. DYE ANALYSIS SCHEDULE
8856-A(01)
Date:
Name(s):
LNB: page no..
Plant(s):
Samples Analyzed:
Atr Filters:
Bulk Dyes:
Spiked Filter Controls:
Extraction/dilution schedule
Time
8:00
:15
'in
:45
9:00
:15
:30
:45
10:00
:15
:30
:45
li-nn
•is
:30
•4S
1:00
"•IS
•10
:45
?-nn
:15
:30
:45
3:001
:1S
:30
:4S
4:00
:lb
:30
:45
5:00
Plant
1.0.
Sample 1.0.
.
le D11n.a
Analysis schedule
Time
8:00
:15
:30
:45
9:00
:15
:30
:45
10:00
:15
:30
:45
11:00
:15
:30
:45
1.00
:1S
:30
:45
2:00
:15
:30
:45
|| 3:00|
:15
:30
:45
4:00
II :15|
II : 30 |
:45
5:00
Sample 1.0.
Cycle
no.
2° 01 In.
ATOT
Corrected
ATOT
D»
00
'Sonicated all PASs and bulk dyes for 10 nln after primary dilution (1° diln.). Agitated all filters on shaker for 30 min
after primary dilution.
Corrected AT(JT = ATQT field air filter sample extract - AT(JT field air filter blank extract.
-------
Form A-2
8856-A(01)
INDIVIDUAL DYE SOLUTION PREPARATION FORM2
BL ion o.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-ll of 20
Plant ID: Prepared by: Date:
Analytical Balance
Calibration Weight
Instrument Calibra
• ID: Model No.: MRI No.:
.s Used: MRI No.:
ition: Post-Tare
Pre-Tare
mg weight = (accuracy = %)
Sample ID No.
'
Pre-Tare (g)
Post-Tare (g)
Dye Weight (g)
aSee Dye Analysis Schedule for dilutions of the above weighed dyes.
A-87
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-12 of 20
11.2 Based on the dye usage data from the drug room site, determine
the major use dye subset (i.e., those dyes which comprise at
least 80% of the total dye quantity handled during the monitoring
period). Using Section II of Form A-3, calculate the relative
weight fraction for each of these dyes. Using this value and the
target dye concentration value (calculated in Step 11.1), calculate
target concentrations of each dye in the dye subset.
11.3 Using Form A-4 and the dye solvent, prepare the spiking standard
solution containing all of these dyes in proportion to their weight
fractions. The total dye concentration of the spiking standard
solution should be as close to the target value (calculated in
Step 11.1) as possible.
11.4 Place a blank air filter (i.e., filter blank from the same lot as
was used at the drug room site) and pad into a three-piece 37-mm
cassette and secure with the extension piece.
11.5 Using a 100-uL syringe, carefully dispense 100 uL of spiking
standard solution from Step 11.3 onto the air filter. The volume
delivered represents the 130% high level of total dyes calculated
in Step 11.1. Deliver the spike in a scattered pattern over the
entire filter. Set aside to dry.
11.6 Repeat Steps 11.4 and 11.5 for the duplicate high-level spiked
filter control.
11.7 Deliver the same volume of spiking standard solution used in
Step 11.5 into an empty 4 oz amber glass jar. Set aside to dry,
leaving the jar uncapped. This is the reference standard at the
high level.
11.8 Repeat Step 11.7 for the duplicate high-level reference standard.
11.9 Repeat Steps 11.4 through 11.6 for the spiked filter controls at
the low dye level, using 50 uL of the spiking standard solution
prepared in Step 11.3.
11.10 Repeat Steps 11.7 and 11.8 for the low-level reference standards
using 50 uL of the spiking standard solution for each.
11.11 Repeat Steps 11.4 through 11.6 for the spiked filter blanks, using
dye solvent in place of the spiking standard solution. Use the
same volume used to spike the high-level spiked filter controls,
e.g., 100 pi for each blank.
11.12 If the pH 3 dye solvent was used in Step 9.2, use the pH 3 dye
solvent (Step 6.4) to perform Steps 11.3-11.11.
A-88
-------
Form A-3
8856-A(01)
SPIKING STANDARD SOLUTION ESTIMATION FORM
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-13 of 20
Plant:
Date:
Name:
I. Total Dye Estimate on Personal Air Filters
Filter ID No.
Corrected
ATOT
Total Dye Estimate (ug)a
Average Total
Dye Estimate
(ug)
awhere ug = A™T x Dilution Volume§ T =
II. Target Values for Dye Subset Comprising £ 80-90% of Total Quantity Dye
Handled:
Spiking Standard Solution's Target Total Dye Concentration =
130% x ug (avg. total dye estimate) _
100 uL (high level spike volume)' ~
Bulk Dye
ID No.
% of Total Quantity
of Dye Handled
% Total
in Dye
Subset =
Relative Weight Fraction =
% Total for Dye
X Total in Dye Subset
Target Cone. (ug/yL)
(Rel. Wt. Fraction x
Target Tot. Dye Cone.)
A-89
-------
Section No.: Appendix B
Revision No.: 3
. . . Date: April 30, 1987
Form A-4 page B_14 of 2Q
8856-A(01)
SPIKING STANDARD SOLUTION PREPARATION FORM AND DATA ON SPIKED FILTERS
Plant: Date:
Analyst: LNB: page no.:
I. Spiking Standard Solution Preparation:
Analytical balance ID:
Calibration weight ID:
Calibration:
Post-Tare: •
Pre-Tare:
mg weight = (accuracy = %)
II. Dye Weights (Use additional sheets if necessary)
Dilution Volume:
Post-Tare: Post-Tare: Post-Tare:
Pre-Tare: Pre-Tare: Pre-Tare:
Dye No. wt.: Dye No. wt.: Dye No. wt.:
Actual Cone.: M9/ML Actual Cone.: ug/pL Actual Cone.: pg/uL
Post-Tare: Post-Tare: Post-Tare:
Pre-Tare: Pre-Tare: Pre-Tare:
Dye No. wt.: Dye No. wt.: Dye No. wt.:
Actual Cone.: ug/uL Actual Cone.: pg/pL Actual Cone.: ug/uL
III. Spiked Filter Data
A. Filter ID:
No. of High-Level Spiked Filter Controls Prepared ; Vol. Spiked uL
No. of Low-Level Spiked Filter Controls Prepared ; Vol. Spiked uL
No. of Spiked Filter Blanks (DMSO only) Prepared ; Vol. DMSO used uL
No. bf "Dry" High-Level Spiking Std. Controls Prepared ; Vol. Spiked uL
No. of "Dry" Low-Level Spiking Std. Controls Prepared ; Vol. Spiked uL
B. Date filters placed on sampling pumps: by (name)
Date filters and "dry" controls are to
be extracted/analyzed: short-term
long-term
longest-term
No. of filters to be prepared/analyzed:
Filters "Dry" Controls
(short-term) high; low; blank high; low
(long-term) high; low; blank high; low
(longest-term) high; low; blank high; low
A-90
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-15 of 20
11.13 Very low dye spiking levels (e.g., < 50 ug) result in weak ab-
sorbances which are difficult to quantitate with a high degree of
confidence. As a result, recovery data based on low spike levels
are more uncertain than those obtained for higher dye spiking
levels. To address this potential problem, low personal air
filter dye estimates will be bracketed by the high and low spiking
levels, but spiking will also be done at an elevated level (eg
100 ug) as well. 'y''
12'° EXPOSURE OF SPIKED FILTER CONTROLS. BLANKS. AND REFERENCE STANDARDS
12.1 Calibrate personal air pumps to draw air at a rate of 2 to
2.5 L/min.
12.2 Attach the dry spiked filter cassettes prepared in Section 11 0
to individual personal air pumps and draw air through the cassettes
for 6 to 8 h. (Relative humidity in the immediate vicinity should
be in the range of 30 to 70%. Fluorescent lights should be on
during the exposure period.)
12.3 During Step 12.2, place the reference standards in the immediate
vicinity of the aerating filters so they are exposed to the same
lighting and humidity conditions.
13.0 ANALYSIS OF DYE SOLUTIONS
13.1 Spectrophotometer Operating Parameters
13.1.1 Turn main power switch to "on" position. Allow Visible-UV
light source to warm up for at least 20 min.
13.1.2 Adjust the following parameters to their proper settings:
Source select: auto
Beam interchange: normal
Mode: autogain
Timer node: off
Slit: 2.0 nm
Abs. zero suppression: 0
Range: 1.0 absorbance units full scale
Chart: off
% zero suppression: off
Period: 0.5 s
Pen: on
Autobaseline: on
A-91
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-16 of 20
Scanning rate: 1 nm/s
Chart display: 50
Scan: off
Function: off
Measurement: serial
Log A offset: 0.1
Concentration: turn knob completely to left
Analog signal output: 0-10 V
13.1.3 Set the upper wavelength limit (e.g., 800 nmb) using the
wavelength 1 dial. Set the lower wavelength limit (e.g.,
330 nm ) using the wavelength 2 dial.
13.1.4 Using the "scan" dial, turn dial (+) or (-) to set
the wavelength at the upper limit (e.g., 800 nm).
13.2 Analog/Digital Computer Interface Box Operating Parameters
Maximum Input Voltage: 10 V
Run Time: interface box run time must exceed actual analysis
run time required to scan the wavelength range, e.g.,
at 1 nm/s from 800 nm to 330 nm, the analysis run
time is 7.83 min. Therefore, an interface run time
of > 8 min is required.
Sampling Time: 1 point/s
13.3 Absorbance Measurement of Filter Extract Solutions
13.3.1 Zero the spectrophotometer by placing the dye solvent
(filtered using a 0.45 urn Gelman Acrodisc) in
both the sample and reference cuvettes. Turn the auto- •
baseline knob to "record" and check to see that red light
comes on. (Autobaseline knob will automatically return to
"on" position.) Adjust the balance knob to give an absor-
bance reading of •* 0.1000 on the digital display. Turn
timer mode knob to "sample and wavelength," then back to
"wavelength." Depress "step" button and scan the wave-
length range (note: upon completion of the scan, the upper
wavelength will automatically be reset). After the wave-
length is reset, turn the timer mode knob to the "off"
position. Adjust the balance knob to give an absorbance
reading of *> 0.0500 on the digital display. Turn timer
aThe maximum voltage must correspond to the maximum signal input of the
A/D box.
b This setting can be anywhere in the range 750-800 nm.
c This setting can be anywhere in the range 300-350 nm. It is usually deter-
mined by the point where the grating in the spectrophotometer changes
from the visible to the ultraviolet region.
A-92
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-17 of 20
mode knob to "sample and wavelength," then back to "wave-
length." Depress "step" button, simultaneously starting
the Nelson interface box, and scan the wavelength range
to obtain a flat baseline from 800-330 nm. Empty the
sample cuvette.
13.3.2 Analyze each filter extract solution, according to the
Dye Analysis Schedule, by withdrawing the filter extract
solution into a 10-mL disposable syringe. Attach a
0.45 \im Gelman Acrodisc filter onto the end of the
syringe and filter the extract solution into the sample
cuvette. Retain the remainder of the extract solution.
13.3.3 Scan the filter extract solution over the wavelength
range, starting the A/D box at the beginning of the scan.
Examine the scan during the analysis run. If the dye
absorjance exceeds 1.0, abort the analysis run immedi-
ately and make an appropriate dilution of an aliquot of
the extract solution. Scan immediately. Document dilu-
tion and time deviations on the Dye Analysis Schedule.
13.4 Absorbance Measurement of Individual Dye Solutions
13.4.1 Zero the spectrophotometer according to the procedure out-
lined in Step 13.3.1.
13.4.2 Analyze each dye solution, according to the Dye Analysis
Schedule, by withdrawing the dye solution from Step 10.3
into a 10-mL disposable syringe, attaching a 0.45 urn
Gelman Acrodisc filter onto the end, and filtering into
the sample cuvette.
13.4.3 Scan the dye solution over the same wavelength range used
for the filter extract solutions, starting the A/D box at
the beginning of the scan.
13.4.4 The maximum absorbance in the dye spectrum should be in
the range of 0.1 to 1.0 absorbance units. If the dye
absorbance does not fall within this range, abort the
analysis run immediately and repeat Step 13.4.3 using an
appropriate dilution of an aliquot from the flask in Step
10.2. Scan immediately. Document dilution and time devi-
ations on the Dye Analysis Schedule.
Turn the timer mode knob to "off" and stop the Nelson Box. Turn the scan
knob to (+) side and increase scan rate until the wavelength is *• 790 nm.
Return scan rate to 1.0 nm/s. Turn scan knob to "off" position at 800 nm.
A-93
-------
Section No.: Appendix B
Revision No.: 3
Date: April.30, 1987
Page B-18 of 20
14-0 ANALYSIS OF SPIKED FILTER CONTROLS AND BLANKS
14.1 Extract all spiked filter controls and blanks as outlined in
bteps 9. 1~9. 6.
14'2 thSughC13S313ed flUer 6XtraCt So1ution as outll'"ed in Steps 13.3.1
14'3 UaCh Peference standard W1'th 8.0 ml of dye solvent, cap, and
14-4 en?? !a?Mefeperce standard solution in the same manner as the
spiked filter extracts in Step 14.2.
15-° Determination of Total Area Counts in a Dye Scan
15.1 For spectrum integration purposes, calculate, to the the nearest
0.01 rain the time required to scan the sample over the desired
wavelength range (Scanning from 800-380 nm at 1 nm/s takes
^tU s = /. 00 mm. )
15.2 Plot the data file in the "re-detect" mode of the Nelson integra-
tion software. Using the cursor, manually integrate the area
above the baseline from 50 s (0.8333 min = 750 nm) to the calcu-
lated time (e.g., 7.00 min = 380 nm).
Note: The height of the baseline is determined by integration of
the dye solvent over the calculated time period. In some cases
due to baseline drift, the baseline used for integrating data
files may deviate slightly. y
15.3 Obtain a hardcopy of the file integration for archiving purposes.
16.0 CALCULATIONS
16 -, Total Absorbance (ATnT) T. . .
•-1 t s A. — T9T •' Tne integration software from Nelson
Analytical divides the maximum signal into 106 parts (counts)
. Since the full scale signal at any particular wavelength is
equal to 1.0 absorbance units, there are 106 area counts per
absorbance unit. Therefore, by dividing the total area counts
va 1 U1 1 I obtained10" 15'0) by 3 faCt°r °f 106> a t0tal absorbance
T values for field ai> filter """P1" and air sampling
are corrected for background absorbance by subtract-
ing the AIQT value for the corresponding field air filter blank(s).
JlQI v?1u65 for individual bulk dye solutions are not subject to
tnese background absorbances and therefore need no correction.
A-94
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-19 of 20
Spectral Absorptivity Constant (a.) for Individual Dyes: I
on Beer's law, each individual dyi^s spectral absorptivity
stant (a ) is calculated using the formula:
con-
where b is the cuvette pathlength in cm (e.g., 1 cm), c is the
dye concentration in ug/mL, and A,nT is the total absorbance of
the dye solution. The dye concentration will be expressed in
terms of a commercial dye basis (i.e., assume 100% purity) and an
active ingredient basic (i.e, using dye purity values supplied by
C.InU/•
16-3 Weighted Average Spectral Absorptivity Constant (a.) for the Dve
Analysis Set:This value is obtained from the individual dye a
constants in the dye analysis set, using the dye handling inforfta-
tion to obtain the appropriate weight fraction for each component
dye.
16.4 Dye Recovery Determination for Spiked Filter Controls:
The recovery value for the group of major use dyes is calculated
from the A,-, values obtained from the spiked filter controls and
the corresponding reference standard solutions. Percent recovery,
R, is calculated using the formula:
R = ATQT of Sp1'ked Fi1ter Control - ATQT of Spiked Filter Blank x 1QQ
of Reference Standard
16-5 Total Dye Estimate on the Air Filter: This is calculated using
the formula:
"TOT ^ •
Total Dye Estimate (ug) = —^1
ag x b
where V is the volume of the filter extract solution in ml (ATn,,~
as, and b are defined above). IUI
16.6 Correction of Dye Estimate for Percent Active Ingredient of Dyes:
The total dye estimate is also reported in terms of the active
ingredient content from the individual dyes. This correction, made
by using dye lot purity information provided by ETAD, is per-
formed on all field air filters as well as their corresponding
cassettes.
A-95
-------
Section No.: Appendix B
Revision No.: 3
Date: April 30, 1987
Page B-20 of 20
16-7 Correction of Dye Estimate for Extraction Efficiency: The total
dye estimate is also corrected for the extraction efficiency of
the analytical method. The overall average dye recovery value is
used to make the correction. This correction is performed only
on the field air filter samples.
16.8 Average Airborne Dye Concentration: This is calculated by divid-
ing the total dye estimate (in mg) by the total volume of air
sampled during the monitoring period (in m3).
A-96
-------
APPENDIX B
DATA QUALITY OBJECTIVES
B-l
-------
DATA QUALITY OBJECTIVES
FOR THE TEXTILE DYE DRUG ROOM STUDY
STUDY OBJECTIVES
Estimate the distribution of the 30 breathing zone dye
concentrations observed in the monitored plants (8 hour time-
weighted concentration), specifically:
a. the average breathing zone dye concentration at the
plants
b. the upper percentiles (85th percentile)
c. confidence intervals for both estimates
2. Determine if dye concentration in breathing zone, adjusted
for time in drug room is correlated with amount of dye
weighed, number of weighings/shifts, or other factors (see
Appendix A), and if BO, determine a functional relationship
between concentration and factors.
3. Estimate the average, and distribution of, amounts of
individual dye compound weighed out during a shift (averages
and histograms need to be presented by dye class and
aggregated)
4. Summarize selected drug room observations and general plant
information in table form (see Appendix B for items to be
summarized)
5. Obtain an extensive first-hand, qualitative view of drug room
operations (the final result will be an individual industrial
hygiene report for each of the 30 plants visited)
Objective 1 is the most important; items 2-5, all of roughly equal
importance, are supplemental to objective 1 but will also enhance
the knowledge of dye exposure. Besides the objectives of the
study, other aspects of the data will be explored (see Appendix
C).
II. BACKGROUND
About 1000 domestic textile dyeing or printing sites have
been identified where there is potential for workers to be exposed
to numerous powder dyes via inhalation or dermal routes. There
are reasonable indications that some dyes or their metabolites
B-2
-------
night be carcinogens or mutagens. Data which document potential
exposure levels of workers associated with the weighing or mixing
cf powder dyes are limited, some of which were obtained from non-
textile dyeing operations. The objective here is to conduct a
well designed study of textile dye weighing rooms in order to
improve the assessment of exposure (and associated risk) with the
use of powder dyes in the American textile industry. This study
is being sponsored jointly by the American Textile Manufacturers
Institute (ATMI), the Ecological and Toxicological Association of
the Dyestuffs Manufacturing Industry (ETAD) and the Office of
Toxic Substances (OTS) of EPA.
III. DATA COLLECTION APPROACH
A two phase approach was chosen for the study. The first
phase was a mailed out questionnaire to 240 plants selected as a
simple random sample from 1390 textile facilities thought to
potentially use powder dyes. The first phase goals were: to make
sure that the list of 1390 plants was not missing any major groups
of dyeing plants (and thus provide a pool from which to draw that
is not grossly unrepresentative); to provide valuable general
information on drug room operations including data useful to
industrial hygienists in preparing for in-plant monitoring; and
possibly to provide assistance for stratified random sampling.
The second phase will be a stratified random sample with the
two strata being: 1. Respondents to first phase questionnaire; 2.
Non-respondents to first phase questionnaire. Plants will be
selected with representation proportional to strata sizes. Actual
in-plant airborne dye-level monitoring and drug room observation
will take place in 30 plants at this phase. The two strata were
chosen to separate first phase respondents from nonrespondents
since there might be differences in the nature of exposure levels,
and since it is desired that the final representation from the two
groups reflect the actual population of dyeing plants.
Within each of the 30 plants selected, a two member team will
record measurements and observations in the plant to satisfy the
study objectives (see Appendices A and B for the approximately 40
items recorded and the Quality Assurance Project Plan for the
methods). For one randomly selected dye weigher at each plant, a
more extensive examination of practices and potential exposure
will be .conducted (only one weigher will be chosen to balance the
needs of the six objectives between themselves in the presence of
limited resources). This weigher will be observed and monitored
during one randomly selected shift. The monitoring will take
place during nearly an entire 8-hour period by personal monitoring
equipment designed to collect solids from the air in the breathing
zone of the selected worker. These will be laboratory analyzed to
determine total dust mass and total dye mass. For the monitored
weigher, observers will also record the amount of each powder dye
B-3
-------
and chemical weighed, the total number of powder dyes weighed, the
amount of time the weigher spends in the weighing area, and other
qualitative and quantitative measurements. Also, the observers
will record information on size of dyeing operation, cleanliness,
ventilation, possible routes of exposure, and other qualitative
information.
IV. DATA QUALITY OBJECTIVES
1. Sources of Variability
There are four major sources of variability anticipated in
the textile dye study: 1. variability in exposure level between
plants; 2. variability between times at the same plant; 3.
variability between weighers (and drug rooms) within the same
plant; 4. chemical analysis error. Due to cost reasons, one
weigher will be randomly sampled to represent weighers within the
plants and the assumption will be made that the dye monitoring
will be conducted at a random time in the plant's operating
schedule (i.e., unrelated to dye level). Sources (1) and (4) are
discussed below.
a. Error Associated with Chemical Analytical Procedures
The 95% confidence bounds for the measured dye level
will differ from plant to plant (since essentially a
unique material will be collected at each plant).
However, the expected 95% confidence bounds will be ?? +
90%* ?? of the reported value (including all error from
the point at which a filter is shipped from a monitored
plant to the final reported value).
There are several components which make up this total
error. For example, the accuracy, or percent recovery,
ranges from ?? 55-75%* ?? (an adjustment will be made
for this, and is included in the total error estimate of
the above paragraph). The error anticipated in using an
average As measurement in the chemical analysis is *
. The precision, which includes instrumental
fluctuations, is _* . Any systematic errors that
may exist, such as operator error in properly preparing
a standard spiked filter used as a control, could not be
accounted for in the precision estimate.
* to be provided by MRI
B-4
-------
b. Sampling Error Associated with the Survey
The sampling error discussed here is the error
introduced from selecting a sample of textile dyeing
plants versus monitoring one shift for the entire
population of textile dyeing plants.
The total sampling error introduced will depend upon the
variability between plants. However, using the data
collected by the National Institute for Occupational
Safety and Health (NIOSH) and the CIBA-GEIGY dye firm,
for three textile dyeing and one paper printing plant,
the standard error expected for estimating the average
exposure is approximately 15% of the estimated average
(2 standard errors = ± 30% on this basis), due to the
sampling error as described. For estimating the 85th
percentile, a nonparametric 95% tolerance interval would
utilize the second highest measured plant value as its
upper bound (again, this includes error associated with
the plant sampling procedure and not such things as
chemical measurement error); however more precise
intervals are likely to be obtained through the use of
probability distributions to estimate the 85th
percentile and its confidence interval.
c. Nonrandom Error Associated with the Survey
There are several potential sources of error (i.e.,
nonsampling errors) besides the four random-type errors,
such as refusal to allow monitoring (this is a voluntary
study), and possibly encountering artificially clean
conditions at the time of the site visit. Only
qualitative, subjective judgments can be made about
these factors. However, to help preserve the usefulness
of the study, an energetically pursued target of
obtaining entry to at least 75% of selected plants has
been set. But given the existing data, a lower bound of
60% monitoring acceptance has been set. Should the
acceptance rate drop below 60% the current objectives
will not be met and a reevaluation of the approach will
be made.
2. Completeness
The survey collection is considered complete when 30 plants
have been monitored, with a second phase response rate of at least
60%, and all 30 chemical analyses are determined to be possible.
B-5
-------
3. Comparability
All chemical analyses will be done at a single laboratory,
Midwest Research Institute. All field collection will be
conducted by Health and Hygiene, Inc. under contract with the dye
industry, and by the firm PEI Associates, Inc. under contract with
EPA. All observational data will be collected via answers to
direct questions or open ended questions with suggested answers
provided, wherever possible, to maintain comparability between
responses at the 30 plants. Also, directing the scope of the
answers on the data collection form insures that the reported
observations actually answer the questions that EPA and the dye
industry want answered.
No other such data collection has been done in the past, nor
has the method for estimating total amount of dye been used
before. The earlier studies by NIOSH and CIBA-GEIGY measured the
amount of a single dye, and only three of the plants were textile
mills.
4. Representativeness (assumptions and universe of interest)
a. The stratified simple random sampling plan will provide
estimates on a national scale and with an
(approximately) known amount of uncertainty due to plant
sampling.
b. Plants were considered to be within the scope of the
study if they dyed or printed textiles using powder dyes
with mechanical equipment. Any amount of dyeing was
considered within the scope of the study.
Drug rooms are considered within the scope of the study
if any amount of weighing takes place there on a regular
basis.
Weighers are the only type of worker represented in the
study. Interest focuses on the weigher since he or she
is the one handling the largest amount of powder dyes.
No direct statement can be made about other workers
except that they are assumed to be exposed to
substantially lower levels of dye.
c. Although the primary goal is to obtain estimates of
average plant levels and the distribution of plant
levels, it is possible to obtain estimates of average
weigher exposure by randomly selecting a worker within
the plant. For example, statements could be made about
the 85th percentile of worker exposure by weighing plant
estimates by the number of weighers in each plant. This
would result in a statement such as "It is estimated
B-6
-------
that 85% of weighers are exposed to levels lower than
xxx mg/m3 during an B-hour shift." Note that this
differs from the estimate of 85th percentile of plant
levels which results in a statement such as: "it is
estimated that the average exposure in 85% of textile
dyeing and printing plants is less than zzz mg/m3, time-
weighted 8-hour average" (xxx and zzz determined from
study).
B-7
-------
APPENDIX A SECONDARY MEASUREMENTS TO EXAMINE FOR ASSOCIATION
WITH DYE LEVEL
The dye concentration adjusted for tine in drug room, (total
dye collected)(tine personal nonitor on/tine in drug room), will
also be examined for correlation with several other variables of
secondary inportance. These are:
1. Production volune of textiles (pounds per year), fron
nailed-out questionnaire
2. Managenent of dye house (vertical, connission or both),
fron nailed-out questionnaire
3. Managenent of dye house (public or private), fron
nailed-out questionnaire
4. Color index class of dyes used, for any dye used during
observed shift by nonitored weigher (acid,
basic/cationic, reactive, direct, disperse, other), fron
site visit log as classified by Chenical Engineering
Branch
5. Total number of dyeing and printing machines serviced by
monitored weigher (average of beginning and end of shift
numbers), fron site visit questionnaire
6. Fiber type dyed or printed (aerylic/nodacrylie,
rayon/cotton, nylon, polyester, other), fron mailed-out
questionnaire
It should be noted that a data set of only 30 observations is
likely to result in some spurious large correlation when many
correlations are calculated. Thus the correlations of this
section will be interpreted in that light.
B-8
-------
APPENDIX B
SUMMARY TABLES
1. ON-SITE QUESTIONNAIRE
Summary tables will be presented for several other variables
collected from the on-site questionnaire with categorized
responses. The variables to be tabled are:
plants by number of weighers, at all shifts during
24 hour period
plants by pounds of dye weighed during shift
plants by number of dyes weighed during shift
plants by number of dye weighings during shift
workers by amount of time in drug room
workers that used dust mask during site visit
workers that used respirator during site visit
workers that smoked in drug room area during site
workers that ate in drug room area during site
1.
2.
3.
4.
5.
6.
7.
8.
9.
Number of
a typical
Number of
Number of
Number of
Number of
Number of
Number of
Number of
visit
Number of
visit
B-9
-------
2. MAILED-OUT QUESTIONNAIRE
As an appendix in the final report, the following variables
will be tabled from the first-phase, mailed-out questionnaire:
11.
12.
@13.
ei4.
§15.
16.
#17.
eis.
ei9.
20.
&21.
&22.
23.
24.
25.
26.
27.
28.
29.
*30.
*31.
32.
33.
34.
Number of textile dyeing plants in each EPA geographical
region
Number of plants by number of dyeing or printing sites within
the company that owns the selected plant
Number of plants by management of house (vertical,
commission, or both)
Number of plants by ownership of house (public, or private)
Distribution of plants by product volume (<1 million
pounds/year, l-8m Ibs/yr, over 8m Ibs/yr)
Number of plants by product line (carpet, yarn, fabric,
other)
Number of plants by type of dyeing or printing equipment
available (batch, semi-continuous/continuous, printing)
Number of plants by fiber dyed or printed
(acrylic/modacrylic, rayon/cotton, nylon, polyester, other)
Number of plants by color index class of powder dye (acid,
basic/cationic, reactive, direct, disperse, other)
Number of plants by number of dyes weighed per 24 hours (less
than 10, 10-19, 20-50, over 50)
Number of plants by pounds of dye used per 24 hours (less
than 100, 100 to 200, 20 - 500, over 500)
Number of plants by number of powder dye weighings per 24
hours (less than 20, 20 - 50, 51 - 100, 101 - 200,
Number of plants by number of dye weighing rooms (
or more rooms)
Number of plants by number of worker shifts per 24 hours
over 200)
1 room, 2
(1,
2, 3)
Number of plants by number of all types of workers exposed at
plant (less than 3, 3, over 3)
Number of plants by number of operating days per week (1 to
4, 5, 6 or 7)
Number of plants by number of employees exposed to powder
dyes (1, 2, 3, 4 or more)
Number of plants for which general ventilation available and
number for which it is not
Number of plants for which a local exhaust or hood is
available and number for which it is not
Number of plants for which dust masks are mandatory and
number for which they are not
Number of plants for which respirators are mandatory and
number for which they are not
Number of plants for which long sleeve clothing is mandatory
and number for which it is not
Number of plants for which impervious gloves are mandatory
and number for which it is not
Number of plants for which goggles are mandatory and number
for which they are not
Notes
Similar information is also tabulated from site visit data
Similar information is also collected in the on-site
questionnaire and is used in the secondary correlation
analysis (see Appendix A)
A portion of the data on this variable will be used in the
secondary correlation analysis (see Appendix A)
B-10
-------
APPENDIX C KOTES ON STUDY OBJECTIVE 1
Although the primary objective of the study is as stated in
objective 1, the approach chosen warrants discussion.
1. The main objective and procedure
One worker (dye weigher) will be monitored at each of the 30
plants visited. This measured value will be used to represent a
typical measurement from that plant. These 30 breathing zone dye
concentrations will be used for an estimate of the distribution of
dye concentrations typically found in plants across the U.S.
2- Another wav of looking at exposure distribution which will be
presented in a report appendix
Note that the above procedure focuses on levels typically
found in textile dyeing plants. Another approach is to look at
the distribution of weigher (worker) exposures. This implies that
a plant with three (3) workers should receive three (3) times as
much weight as a plant with one (1) worker, since the workers
there represent three (3) times as much of the population. This
approach has considerable appeal; however, practical problems are
discussed in Section 4 of this Appendix.
3. Distinction with the main ob-ieetive
Although Section 2 of this Appendix discusses the main
difference between the chosen procedure focusing on typical levels
found in plants and the alternative approach, this section
elaborates upon the distinction. •
In contrast with objective 1, the alternate approach takes
into account the number of weighers working at each plant and is
weighted according to the number working at the monitored plant
during a typical 24-hour period. For example, statements could be
made about the 85th percentile of worker breathing zone dye
concentration by weighing plant estimates by the number of
weighers in each plant. This would result in a statement such as
"It is estimated that 85% of weighers are potentially exposed to
concentrations lower than xxx mg/m3 during an 8-hour shift." Note
that this differs from the estimate of the 85th percentile of
plant levels (objective 1) which results in a statement such as:
"It is estimated that the average concentration in 85 percent of
textile dyeing and printing plants is less than zzz mg/m3, time-
weighted 8-hour average" (xxx and zzz determined from study).
B-ll
-------
4. Reason for chosen emphasis
Although it would be ideal to sample every weigher at every
shift of the sampled plants, it is not feasible for several
reasons:
o cost; monitoring more than one weigher per visit would
raise laboratory and site visit costs dramatically. At
least one extra person would need to travel to the plant
site to record the dyes weighed by an extra weigher, the
relative amounts of each dye weighed, set up the
personal air samplers, and collect specimens of each dye
weighed by the additional weigher. The laboratory would
then need to develop a new procedure to chemically
analyze the mixture of dyes weighed by the second
worker.
o personnel maximum; monitoring workers by having more
than two (2) observers per site is often not feasible
due to limited workspace at the sites. Furthermore,
cooperation in this voluntary study could be impaired by
requesting to send a larger delegation.
o balance with other objectives; sending a larger team to
each site would allow for some information on the
importance between worker variability and allow for a
somewhat better estimate of the distribution of
breathing zone dye concentrations associated with the
population of dye weighers. However, much of the study
concentrates on observations made on conditions and
possible routes of exposure at the plant/drug room
visited (i.e., more dye measurements would not
necessarily improve these aspects of the study
appreciably); the potential routes of exposure on a
plant by plant (or drug room by drug room) basis are
also of direct Interest.
o correlation between reported dve concentrations at the
same plant: simply monitoring two workers at the same
plant does not provide for two independent estimates of
dye concentration due to the nature of the chemical
analysis method. That is, the error in the chemical
measurement is such that repeated measurements at the
. same site will tend to be repeatedly too high or too low
if the same dyes tend to be used (and, less importantly,
that the mechanism for exposure remains the same). This
is due to the fact that the proportional amount of dyes
in the mixture on the filter is not known, so that the
error in using the average absorbance value tends to be
in the same direction for repeated chemical analyses of
similar mixtures (also, the percent recovery estimate
will have a similar tendency to be repeatedly too high
or too low for the same unknown mixture). However, this
should cancel out when estimating the 30-plant average.
B-12
-------
APPENDIX C
LETTERS SENT WITH QUESTIONNAIRE
C-l
-------
a
AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC
MOI CONNECTICUT AVENUE. NW. SUITE 300. WASHINGTON. DC 20036
TWX 710-822-9489
TEL 202 862-0500
December 20, 1985
Dear
This confirms our recent telephone conversation relative to the
Textile Drug Room Monitoring Study. Your participation In this survey
is of the utmost importance. The answers you provide will help us move
to the second phase of a survey whose ultimate goal is to gather
information that will be shared with the U.S. Environmental Protection
Agency (EPA) to characterize the workplace exposure levels of the dyes
in dye weighing and mixing rooms of textile dyehouse and printing
operations.•
The value for an exposure level which EPA currently uses in
existing chemical assessments is based on the level detected in the
leather industry, which is probably a higher number than might be
observed in the textile industry. Ultimately, in this survey, the
measured levels of dye in the air of 30 drug" rooms of the sampled
textile dyeing and printing plants will provide a more representative,
and most likely a lower, value than the one the EPA is currently using.
This survey has the support of ATMI and also the Dyesruff Manufacturers.
The cor^uJsTbns drawn from the current survey will'affect all members of
the textile and dye industry.
We look forward to your completed questionnaire. Please feel free
to call Maggie Dean of ATMI at (202) 862-0^0 if you have any questions.
Sincerely,
Carlos Moore
Executive Vice President
C-2
-------
we have included with the survey forms, instructions and an example of a
completed survey. If any proprietary business information is provided which
should be treated as confidential, so indicate by encircling the Y on the
confidential business information (CBI) line at the top of the appropriate
page(s). Your request for confidentiality will be respected. We would
appreciate hearing from you by December 5. 1985. If you have any questions
please contact Maggie Dean (202) 862-0580. We extend our thanks for your
cooperation and assistance in this industry-government cooperative program.
Sincerely,
Carlos Moore
Executive Vice President
Enclosure
OIS/7T/1-2/CS
C-3
-------
D
AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC.
1101 CONNECTICTT AVEM'E. N U . SUITE 300 WASHINGTON. D C 20036
TWX 7IO-822-W89
TEL 202 862-0500
May 23, 1986
Thank you for participating in the first phase of the joint study with AM,
the Ecological and lexicological Association of the Eyestuffs Manufacturine
Industry (ETAD) and the U.S. Environmental Protection Agency (EPA) to
determine worker exposure to dye dust. The second phase of this project
Inwlves the measurement of actual exposure levels in a representative group
of U.S. textile plants. Your plant is one of thirty textile plants selected
at random for the second part of this important study. The validity of the '
study depends upon achieving a high level of participation from the 30
plants. We encourage your participation for this reason.
The next phase will be conducted on-site by a field team of two certified
industrial hygienists. Total time on-site over a 2-day period will be about
14-16 hours. Every effort will be made to minimize any inconvenience to your
plant from the visit. In fact, the validity of the study can only be
maintained by avoiding disruption of workplace activities. An individual site
report will be prepared and you will receive a copy following completion of
the study. The overall survey results will be presented in a 30-site
composite report, individual plants will not be identified in this final
report.
More details of the study are provided in the attached statement on objectives
and protocol. •
C-4
-------
fhe American Textile Manufacturers Institute. Inc. (ATMI) and the Ecological
and lexicological Association of the Dyestuffs Manufacturing Industry (ETAD)
arr performing a study to determine the extent to which textile workers may be
exposed to dye dust when weighing and mixing powder dyes. This study is being
•lone in cooperation with the US Environmental Protection Agency (EPA).
You. plant is one of 200 textile dyeing or printing sites which has been
randomly selected for surveying. The purpose of this preliminary survey is to
gather information about production capabilities, fibers processed, product
lines, dyeing/printing operations and prevailing practices of worker
protection in order to obtain a broad picture on the use of dyestuffs for
textiles and the potential for worker exposure. From these 200 companies 30
will be selected as representing typical plants in the textile dyeing/printing
Industry. They will be requested to permit voluntary monitoring using
personal monitors of the workers in drugrooms by a private consulting firm.
Health and Hygiene Co. of Greensboro. N.C. who will be accompanied by an EPA
representative. The professional staff of Health and Hygiene have
considerable textile experience and are highly suited and well qualified for
tins assignment. Monitoring data will be used by Industry and EPA as a basis
for estimating potential levels of exposure of workers to powder dyes No
regulatory action will be generated against Individual participants as a
re!U!L°! th,1s data 9«ther1ng program. Data will be coded by ATMI and
submitted only in coded form to EPA. After the survey forms have been
reviewed for completeness, the code will be destroyed.
Jn!UeSt!? t0 !?1!!BJirnjr comPlete the enclosed survey forms which will
information which Is necessary to Initiate the program. To aid you,
C-5
-------
There will be some direct benefits to participants in this study, they include:
1. A confidential report vhich characterizes industrial hygiene
practices at the plant and measures dye concentration levels in
the drug room.
2. An opportunity for participating companies to ccnpare their
results with those of other participating plants.
3. The current EPA approach, which is based on worst case
assumptions, is excluding from the U S. market many new dyes
which are available to U S. competitors abroad. A direct
objective of this study is to secure a more realistic assessment
of potential risks.
4. A ftnal report which characterizes the industry based on
consolidation of results from the questionaire and monitoring
study of all participants. This report will not identify
specific sites.
Within the next few days Ms. Maggie Dean, ATMI's Director of Safety, Health
and Environment, will phone you to discuss arrangements and possible dates for
the two-day on-site visit by the field study team. Currently, we plan to
begin site visits the week of June 16. Shortly after, Dr. William Dyson will
phone you to schedule the on-site visit and answer any specific questions
about the monitoring procedure. Your participation in this second and final
phase of the study is important and I personally encourage you to take part.
If you have any questions, please contact Maggie at (202) 862-0380.
Sincerely,
Carlos Moore
Executive Vice President
CM:jt
C-6
-------
TEXTILE CRUG ROOM MONITORDC STOP? (TERMS)
OBJECTIVES AND MECHANISM
About 1000 domestic textile dyeing or printing sites have been Identified
where there is potential for workers to be exposed to numerous powder dyes via
inhalation or dermal routes. There are reasonable indications that some dyes
(both new submissions and existing) or their metabolites might be carcinogens
or mutagens. Data which document potential exposure levels of workers
associated with the weighing or mixing of powder dyes are limited, some being
derived from other than the textile industry. Our objective is to conduct a
well-designed study of textile dye weighing rooms in order to improve the
assessment of. exposure (and associated risk) with the use of powder dyes in
the American textile industry. This study is being sponsored jointly by the
American Textile Manufacturers Institute (AIM), the Ecological and
lexicological Association of the Dyestuffs Manufacturing Industry (ETAD) and
the Office of Toxic Substances (OTS) of EPA.
The mechanism of accomplishment will consist of monitoring one randomly
selected dye weigher in each of 30 randomly selected sites which use solid
dyestuffs in the dyeing or printing of textile fibers. The on-site field team
will consist of two certified industrial hygienists. One will be an employee
of Health and Hygiene. Inc. (H4H) of Greensboro, NC. The other will be an
employee of PEI Associates, Inc. (PEI) of Cincinnati, OH.
C-7
-------
to site obteivers will record trt i^ncity oi cacii soiixi substance weighed by
the person being monitored, the number of weights of each, the total mass of
each substance weighed and the duration of time that the worker is within the
weighing area. They will characterize each work area in respect to muter of
dyeing/printing units in operation, materials flow, cleanliness and
ventilation including engineering controls. They will also characterize each
person being monitored relative to work habits, work history and protection
equipment utilized including clothing and personal protective controls.
Over an 8 hour period, solids will be collected from the air in the breathing
zone of each selected worker. At a later time, collection filters will be
analyzed for total dust and for total dye. Observers will also collect
analytical standardization controls consisting of samples of each dye and
other materials which may interfere with the analysis.
The field team is expected to be at each site for a period no greater than 2
consecutive days. Projected activities and duration will be as follows.
Activities Duration
Pre-Monitoring Gain site familiarity 2-U hrs
Characterize site
Monitoring Monitor worker 8 hrs
Record Dyes/chemicals
weighed
Characterized worker
C-8
-------
Post-monitoring Collect samples 2-4 hrs
Review and completion
of data/sample collection
Review with plant offical
Observers will preserve the confidentiality of operations and formulations of
the sites visited. As a courtesy and to ensure validity, they will
conscientiously refrain from any interference in workers' performance of
duties. Within one month of the monitoring date, observers will issue a joint
report which characterizes each site and the nonitored individual.
C-9
-------
AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC.
HOI CONNECTICUT AVENUE. N W. SUITE 300 WASHINGTON. DC 20036
TWX 710-822-9489
TEL 202/862-0500
This letter is in regard to a contact you had recently with WESTAT relative to
a joint industry/EPA sponsored study of textile drug rooms. We want you to
know ATMI is actively supporting this effort and your participation in the
survey is very important. The answers you provide will help us move to the
second phase of the study, which is to assess exposure levels to dyes in
weighing and mixing rooms of textile dyehouse and printing operations. The
information is intended to help EPA in evaluating applications for manufacture
of new dyes. You perhaps are aware that the Toxic Substances Control Act
requires the agency to assess potential health risks before approving the
manufacture of any new chemicals or chemical compounds.
EPA's current strategy for assessing dyes is based on exposure levels found in
the leather industry thereby resulting in fewer approvals of new dyes.
Ultimately, this study will establish levels representative of dye in the air
of 30 textile drug rooms. We expect it to support our position that exposure
levels in the textile industry are much lower than those assumed by EPA.
We urge you to join in this study and we look forward to receiving your
completed questionaire. If you have any questions, please feel free to call
Maggie Dean of ATMI at (202) 862-0580.
Since
Director /"Government
Reglations/Regulatory
US'
010
-------
<• ^
D
AMERICAN TEXTILE MANUFACTURERS INSTITUTE, INC.
1 101 CONNECTICUT AVENUE. NU.. SUITE 300. WASHINGTON DC 3O03t
TWX 7l(«22-»48?
TEL 202/862-0500
January 15, 1987
Dear
•
. ,«. "ft." ""•"•
"" pr°Vided in ^ atta^d statement on objectives
C-ll
-------
There will be some direct benefits to participants in this study; they include:
1. A confidential report which characterizes industrial hygiene
practices at the plant and measures dye concentration levels in
the drug room.
2. An opportunity for participating companies to compare their
results with those of other participating plants.
3. The current EPA approach, which is based on worst case
assumptions, is excluding from the U S. narket many new dyes
which are available to U S. competitors abroad. A direct
objective of this study is to secure a more realistic assessment
of potential risks.
4. A final report which characterizes the industry based on
consolidation of results from the questionaire and axmitoriiig
I*" ific " P81"11"?81*5' ™s "P0" *i" «* identify
iT? this Second "* final P1*5' * the study is very
important to the project's success and we personally encourage you to take
part. We plan to conduct site visits during the next few months and will
'
ttestud'v Sf1^ ? SeCUrC JOUr a*reeDent «* answer "V questions about
tne study and monitoring procedures.
Sincerely,
Maggie Dean
Director
Safety Health and Environment
ATMI
Eric Clarke
Executive Secretary
Ecological and lexicological
Association of the Dyestuffs
Manufacturing Industry
C-12
-------
TEXTILE DRUG ROOM MONITORING STUDY (TDRMS)
OBJECTIVES AND MECHANISM
About 1000 domestic textile dyeing or printing sites have been identified
where there Is potential for workers to be exposed to numerous powder dyes via
inhalation or dermal routes. There are reasonable indications that some dyes
(both new submissions and existing) or their metabolites might be carcinogens
or mutagens. Data which document potential exposure levels of workers
associated with the weighing or mixing of powder dyes are limited, some being
derived from other than the textile industry. Our objective is to conduct a
well-designed study of textile dye weighing rooms in order to improve the
assessment of exposure (and associated risk) with the use of powder dyes in
the American textile industry. This study is being sponsored jointly by the
American Textile Manufacturers Institute (ATMI). the Ecological and
Toxicological Association of the Dyestuffs Manufacturing Industry (ETAD) and
the Office of Toxic Substances (OTS) of EPA.
The oechanism of accomplishment will consist of monitoring one randomly
selected dye weigher in each of 30 randomly selected sites which use solid
dyestuffs In the dyeing or printing of textile fibers. The on-site field team
will consist of two certified industrial hygienists. One will be an employee
of Health and Hygiene. Inc. (H+H) of Greensboro. NC. The other will be an
employee of PEI Associates. Inc. (PEI) of Cincinnati. OH.
0-13
-------
On site observers will record the identity of each solid substance weighed by
the person being monitored, the number of weighings of each, the total mass of
each substance weighed and the duration of tine that the worker 1s within the
weighing area. They will characterize each work area in respect to number -of
dyeing/printing units in operation, materials flow, cleanliness and
ventilation including engineering controls. They will also characterize each
person being monitored relative to work habits, work history and protection
equipment utilized including clothing and personal protective controls.
Over an 8 hour period, solids will be collected from the air In the breathing
zone of each selected worker. At a later time, collection filters will be
analyzed for total dust and for total dye. Observers will also collect
analytical standardization controls consisting of samples of each dye and
other materials which may Interfere with the analysis.
The field team is expected to be at each site for a period no greater than 2
consecutive days. Projected activities and duration will be as follows:
Activities Duration
Pre-Monitoring Gain site familiarity 2-4 hrs
Characterize site
Monitoring Monitor worker 8 hrs
Record Dyes/chemicals
weighed
Characterized worker
C-14
-------
Post-monitoring Collect samples 2-4 hrs
Review and completion
of data/sample collection
Review with plant offical
Observers will preserve the confidentiality of operations and formulations of
the sites visited. As a courtesy and to ensure validity, they will
conscientiously refrain from any interference in workers' performance of
duties. Within one month of the monitoring date, observers will Issue a joint
report which characterizes each site and the monitored individual.
C-15
-------
June 12, 19B7
Dear :
The American Textile Manufacturers Institute, Inc.. the Ecological and
Toxicological Association of the Dyestuffs Manufacturing Industry and the
Office of Toxic Substances of the U.S. Environmental Protection Agency, thank
you for participating in the joint industry-government textile drug room
monitoring study.
The monitoring phase of the survey of dustiness in textile drug rooms of 24
randomly selected volunteer plant sites was completed in May 1987. Reports
are now being prepared characterizing, but not identifying, each site which
participated. A copy of the report on your facility will be forwar
-------
D
AMERICAN TEXTILE MANUFACTURERS INSTITUTE. INC.
HOI CONNECTICUT AVENUE. N W . SUITE 300 WASHINGTON DC 20036
TW\ 710-822-9489
TEL 202'862-0500
November 18, 1985
Dear
The American Textile Manufacturers Institute (ATMI) and the
Ecological and Toxicological Association of the Dye stuffs Manufacturing
Industry (ETAD) are performing a study to determine the extent to which
textile workers may be exposed to dye dust when weighing and mixing
powder dyes. This study is being done in cooperation with the U.S.
Environmental Protection Agency (EPA) .
Your plant is one of 200 textile dyeing or printing sites which
has been randomly selected for surveying. The purpose of this
preliminary survey is to gather information about production
capabilities, fibers processed, product lines, dyeing/printing
operations and prevailing practices of worker protection in order to
obtain a broad picture on the use of dyestuffs for textiles and the
potential for worker exposure. From these 200 companies, 30 will be
selected as representing typical plants in the textile dyeing/printing
industry. They will be requested to permit voluntary monitoring using
personal monitors of the workers in drugrooms by a private consulting
firm, Health and Hygiene Company of Greensboro, North Carolina, who will
be accompanied by an EPA representative. The professional staff of
Health and Hygiene have considerable textile experience and are highly
suited and well qualified for this assignment.
Monitoring data will be used by industry and EPA as a basis for
estimating potential levels of exposure of workers to powder dyes. No
regulatory action will be generated against individual participants as a
C-17
-------
result of this data gathering program. Data will be coded by ATMI and
review^ fy V^ ^ t0 ^ **M ^ Surve* *°™ Have been
reviewed for completeness, the code will be destroyed.
You are requested to voluntarily complete the enclosed survey
forms which will provide information which is necessary to initiate the
program. To aid you, we have included with the survey forms
instructions and an exaaple of a completed survey. If any proprietary
business information is provided which should be treated as
confidential, so indicate by encircling the Y on the confidential
business information (CBI) line at the top of the appropriate page(s)
lour request will be respected. We would appreciate hearing from you'by
S" •";;«?; L^'ncc" y°U haVC a^ «n»«l«».. Pl«" contact Maggie *
Dean, (202) 862-0580. We extend our thanks for your cooperation and
assistance in this industry-government cooperative program.
Sincerely,
Carlos Moore
Executive Vice President
Enclosure
OIS/7T/1-2/CS
C-18
-------
APPENDIX D
FIRST PHASE QUESTIONNAIRE
D-l
-------
PLEASE GIVE THE ACTUAL PHYSICAL LOCftUQN OF THE FACILITY
RATHER THAN THE MAKING ADDRESS (IF DIFFERENT FROM LABEL).
Street Address:
City: __
State:
STREET OR P.O. BOX
Zip Code: |_L_L_I_I_I
PUASE RECORD THE NAME AND ADDRESS OF THE PARENT COMPANY
(CORPORATE HEADQUARTERS) OF THIS FACILITY. IF THIS FACILITY
HAS NO PARENT COMPANY, PLEASE WRITE 'NOT APPLICABLE' IN THE
SPACE PROVIDED FOR PARENT COMPANY NAME.
Parent Company Name:
Address:
STREET
CITY
STATE
PLEASE RECORD THE NAME, TITLE AND PHONE NUCER OF THE
PERSON MHO MAY BE CONTACTED FOR FATHER INFORMATION.
Name of Contact i
Title:
Phone Ni»ber: ( )_
D-2
-------
GENERAL INSTRUCTIONS
All information requested concerns solid (powder or granular) dyes only. Exclude information
for operations using only pigments or liquid dyes.
Located in the upper right corner of each page of the questionnaire is a box in which you ma>
indicate the presence of confidential business information (CBI) on that page. Pieace circle
Y (yes) or N (no) to indicate the presence or absence of proprietary information on each page.
Most questions will have specific instructions to assist you in their completion. If you have
a situation not covered by the instructions, please don't hesitate to call the number in the
box below for assistance.
An identification number will be asaigned by ATMI as a reference in order to protect the
identity of each participant. This number .will appear in the top right corner of each page.
PLEASE COMPLETE THIS QUESTIONNAIRE AS SOON AS POSSIBLE. AS SOON
AS IT HAS BEEN COMPLETED, RETURN IT TO IN THE ENCLOSED,
POSTAGE-PAID ENVELOPE. IF YOU HAVE ANY QUESTIONS ABOUT HOW. TO
COMPLETE THE QUESTIONNAIRE, PLEASE CALL AT
PLEASE RECORD THE FOLLOWING INFORMATION ONLY IF DIFFERENT FROM LABEL ABOVE.
Name of Dyeing or Printing Facility:
Nailing Address:
STREET OR P.O. BOX
CITY STATE
1 1 1 1 1 1
ZIP
NOVEMBER 1985
D-3
-------
CBI: Y N
1. "hat are the total number of sites within the entire corporation where dyeing or printing
operations are performed? (PLEASE COUNT EACH SEPARATE DYE HOUSE OR PRINT SHOP.)
# OF SITES:
2. Is thie particular facility publicly or privately owned? (NARK AN 'X' IN THE APPROPRIATE
WBLIC | | PRIVATE | |
3. Does this facility operate on a vertical (integrated) or coamission basis? (NARK AN 'X1 IN
THE APPROPRIATE BOX.)
VERTICAL | | COKM1SSION | |
D-A
-------
CBI:
IDf:
PRODUCT LINCS DYED OR PRINTED
a. In the table below, please specify the following information:
COLUMN 1: IN THIS COLUMN PLEASE MARK AN 'X' IN THE BOX CORRESPONDING TO EACH PRODUCT EITHER
DYED OR PRINTED AT THIS FACILITY. PRODUCT LINES WHICH DO NOT FIT ANY OF THE
INDICATED CATEGORIES SHOULD BE ENTERED ON THE OTHER (SPECIFY) LINES.
COLUMN 2: IN THIS COLUMM, INDICATE ON THE APPROPRIATE LINE THE AMOUNT OF EACH PRODUCT WHICH
IS SUBJECTED TO A DYING OPERATION USING POWDER DYES. THIS SHOULD BE POUNDS OF
GOODS PROCESSED. IF IT IS INCONVENIENT FOR YOU TO SPECIFY THE NUMBERTTTDjNDS
PLEASE MARK AN 'X1 IN THIS BOX I | AN'E) RECORD YOUR ANSWER IN PERCENT OF TOTAL
PRODUCTION. THE TOTAL FIGURE AT~THE BOTTOM OF THE COLUMN SHOULD BE GIVEN IN
POUNDS OF FIBER DYED (WITH POWDER DYES) PER YEAR.
COLUMN 3: IN THIS COLUMN, INDICATE ON THE APPROPRIATE LINE THE AMOUNT OF EACH PRODUCT WHICH
IS PRINTED WITH POWDER DYES. THIS SHOULD BE AS POUNDS OF GOODS PROCESSED. IF IT
IS INCONVENIENT FOR YOU TO SPECIFY THE NUMBER OF POUNDS PLEASE MARK AN 'X1 IN
THIS BOA I TXVlD RECORD YOUR ANSWER IN PERCENT OF TOTAL PRODUCTION. THE TOTAL
FIGURE AT IRE BOTTOM OF THE COLUMN SHOULD BE GIVEN IN POUNDS OF FIBER PRINTED
(WITH POWDER DYES) PER YEAR. IF ANY GOODS ARE BOTH DYED AND PRINTED AT THIS
SITE, LIST THE QUANTITIES IN BOTH COLUMN 2 (DYE) AND COLUMN 3 IPRIKI1) BRACKETED
BY PARENTHESES.
PRODUCT LINE
•. Staple
b. Yarn/thread
c. General apparel - woven
d. General apparel - knitted
e. Carpet/rug (include automotive)
f. Other hone furnishings
g. Other transportation fabrics
h. Other pile fabrics
i. Outerwear/cloaking fabrics
j. Towels/terry cloth
k. Sheets/linens
1. Linings/woven continuous filament
at. Marrow fabrics
n. Light wight fabrics
o. Hoaiery/intiaate war
p. Fabricated goods (as sweaters, socks, etc.)
q. Other woven fabrics (SPECIFY)
r. Other knitted fabrics (SPECIFY)
a. Other aubstrataa (SPECIFY)
Total pounds per year j
D-5 |
COLUMN 1
PRODUCT
DYED OR
PRINTED
L_l
L_J
l__l
I I
LJ
I I
l_l
LJ
L_l
I— I
LJ
LJ
l__l
LJ
LJ
l_l
I_J
LJ
l_l
LJ
LJ
I_l
l_l
LJ
l_l
LJ
rra.1
•aHa^aV
COLUMN 2
•MOUNT DYED
PER YEAR
IN POUNDS
1
|
COLUMN 3
AMOUNT
PRINTED
PER YEAR
IN POUNDS
1
-------
CBI: Y N
">«: U_U_U
5.
Which of the following fibers are dyed or printed at this facility? (MARK AN 'X' IN THE BOX
FOR EACH FIBER USED. IF FIBERS OTHER THAN THOSE LISTED ARE USED, PLEASE SPECIFY IN THE
SPACE PROVIDED.)
Acrylic/Modacrylic. . . I I
Rayon | |
Cotton | |
Nylon U
Polyester |_|
Acetate ........ | _ |
Other (SPECIFY)
U
U
6.
Nhich of the color index classes of powder (solid) dyes are used at this facility'' (HARK
AN 'X' IN THE BOX FOR EACH DYE CLASS USED. IF DYE CLASSES OTHER THAN THOSE LISTED ARE USED,
PLEASE SPECIFY IN THE SPACE PROVIDED.)
Acid (include •etallized). . . | |
Basic/Cationic | |
Reactive | |
Direct LI
Disperse | |
Naphthol/Azoic | |
Chrome/Mordant | |
Sulfur |_|
Vet U
Other (SPECIFY)
- - LI
- - LI
D-6
-------
CBI: Y N
IM: III
SPECIFY IN THE SPACE PROVIDED.)
A. Batch
Beck. .
Beam. .
Jet . .
Jigg. .
Package
Paddle.
Rotary
Stock/Top . . .
Skein
Other (SPECIFY)
L
.1
Pad .
Flood
0-Box
Spray
B. Cont inuous /Semi -Cont invjous
| | Warp Dye. ...
I I Other (SPECIFY)
Flat Bed Screen:
Hand
Machine
Rotary Screen . .
Roller
C. Print
Space Dye
Polychromatic . . •
Other (SPECIFY)
D-7
-------
CBI:
FREQUENCY AND VOLUME OF POWDER DYE HANDLING
B. In the table below, please specify the following information:
COLUMN 1: IN THIS COLUMN, PLEASE RECORD YOUR BEST ESTIMATE OF QUANTITIES SPECIFIED FOR A
TYPICAL DAY/WEEK BASED ON PRODUCTION FIGURES OVER THE PAST YEAR.
COLUMN 2: IN THIS COLUMN, PLEASE RECORD YOUR BEST ESTIMATE OF BOTH THE LOW AND HIGH LEVELS
OF QUANTITIES SPECIFIED BASED ON PRODUCTION OVER THE PAST 12 MONTHS. DO NOT
INCLUDE EXTREMES WHICH ARE THE RESULT OF ABNORMAL CONDITIONS SUCH AS ZERO PRODUC-
TION DURING A CHRISTMAS WEEK CLOSING, OR EXCESSIVELY HIGH PRODUCTION RESULTING
FROM A "CATCH-UP WEEK" FOLLOWING AN EXTENDED POWER OUTAGE.
Column 1
Typical
Colian 2
Law/High for Year
a. Number of individual powder dyes weighed per
24 hour day; (INCLUDE EACH DYE ONLY ONCE
REGARDLESS OF THE NUMBER OF TIMES IT IS USED)
b. Number of pounds of powder dyes weighed per
24 hour day;
c. Number of weighings of powder dyes per 24 hour
day; (COUNT EACH DYE WITHIN A FORMULA SEPAR-
ATELY. INCLUDE BOTH STARTING FORMULAS AND
COLOR ADDS OR FEEDS);
d. Number of dye weighing rooms or areas at this
facility;
a. Number of ahifte operating per typical
day/24 hours;
f. Number of daya of operation per typical week;
g. Number of employees Mho weigh, «ix, handle
or are otherwise exposed to powder dyes on a
typical day.
D-8
-------
C6I: V N
DRUG ROOM EXPOSURE CONTROLS WHEN WEIGHING POWDER DYE
9. In the table below, please specify the following information:
COLUMN 1: IN THIS COLUMN, PLEASE MARK AN 'X' IN THE BOX FOR EACH CONTROL IF IT IS AVAILABLE
BUT ITS USE IS NOT REQUIRED.
COLUMN 2: IN THIS COLUMN, PLEASE HARK AN 'X1 FOR EACH CONTROL IF ITS USE IS MANDATORY
CITHER BY COMPANY DECREE OR REGULATION.
GENERAL: "ANY CONTROL NOT MARKED IN EITHER COLUMN MILL BE ASSUMED TO NOT BE IN USE AT THIS
FACILITY. IF CONTROLS OTHER THAN THOSE LISTED ARE USED, PLEASE SPECIFY IN THE
SPACE PROVIDED.
B. General ventilation
b. Local exhaust or hood
c. Dust mask
d. Respirator
e. Long sleeve clothing
f. Inpervious gloves
g. Goggles
Other (SPECIFY)
h.
i.
Colunn 1
Available
tot Required
I— I
W
LJ
LJ
LJ
LJ
LJ
LJ
LJ
Col inn 2
Mandatory Use
I—I
LJ
LJ
LJ
LJ
LJ
LJ
LJ
LJ
D-9
-------
APPENDIX E
DATA TABLES FROM FIRST PHASE QUESTIONNAIRE
E-l
-------
DATA TABLES TO BE PROVIDED IN NEXT DRAFT
E-2
-------
APPENDIX F
IN-PLANT QUESTIONNAIRE
F-l
-------
Plant I.D. (2 digits) /
Recorder:
ATM I
ETAD
US EPA -
JOINT DYE/DRUG ROOMS MONITORING SURVEY
OF TEXTILE DYEING & PRINTING INDUSTRIES BY
American Textile Manufacturers Institute
Ecological & Toxicological Association of the Dyestuffs
Manufacturing Industry
United States Environmental Protection Agency
Name of Dyeing or Printing Facility:
Street Address:
City:
, State:
Zip Code: / / / / / /
Name of Contact:
Title:
Phone Number: ( )
SELECTION OF WORK SHIFT
i A » 2 ?K BhKftS W?ere P°wder dyes are ^ighed. Circle number in
Column A. Based on the above phone number, circle the appropriate digits in
Column B. Circle the observed shift in Column C. »P"«e aigits in
* A
No. of shifts where
powder dye weighed
1
2
3
Middle 3 digits of 7-digit
telephone number
000-999
000-499
500-999
000-333
334-666
667-999
Shift to be observed
1st
1st
2nd
1st
2nd
Record Times of Shift: Prom _ : _ a.m./p.m. To _ : _ a.m. /p.m.
SELECTION OF WEIGHER IN OBSERVED SHIFT TO BE ASSIGNED MONITOR
?flec* the weigher to be monitored from the chosen shift, collect the last 3
*"
three
ciosest *°
SSN:
I.D. f
F-2
Date: / /
•o. date year
-------
Plant I.D. (2 digits) /
Recorder:
SAMPLING AREA INFORMATION
SKETCH LAYOUT OF DRUG ROOM/AREA FOR WEIGHING POWER DYES
'
with
IMPERATURE AND HUMIDITV
end of the
F-3
-------
Plant I.D. (2 digits) /
Recorder: '
WZIGHFR WORKER INTERVIEW
To be asked of weigher (Name) __ selected on first page.
I'd^like to ask some questions about you and your work history handling powder
1. When were you born? / /
month day year
2. How many years have you worked at this site?
3. How many years have you handled (weighed and/or mixed)
powder dyes at this site? „.„
— ^^_^___ years
4. In your lifetime work experience, indicate how long you have
worked in any industry handling (weighing and/or mixing)
powder dyes? yj _ years
5. What information have you received in the safe handling of dyes, such as
training, courses, and literature? _
OBSERVATION OF WEIGHER PRACTICES (Part I)
i T^J If Ves' describe
1. Did monitored weigher
wear dust mask or No // Yes //
respirator? — •
2. Did monitored weigher
smoke in weigh area? No l_ (_ Yes / /
3. Did monitored weigher
eat in weigh area? No / (_ Yes /£
QUESTIONS FOR WEIGHER'S SUPERVISOR
"" lfhaLls th? t°t!1 number of dyeing or printing machines being serviced today
by the monitored weigher: J
Equipment
Number at
Start of Shift
Number at
End of Shift
A. Printing
B. Semi-continuous/Continuous
C. Batch dyeing
What is the total number of weighers at your plant on a typical day?
a. first shift
b. second shift
c. third shift F-4
d. Total
-------
Plant I.D. (2 digits) /
Recorder:
DETAILED WORKER/SAMPLING AREA OBSERVATIONS
Actual Time Periods Monitored: From : a.m./p.m. To : a.m./p.n,
Job Category:
Specific Duties:
A. Clothing & Personal Protective Equipment Utilized:
B. Worker Practices: Note equipment used and dermal contact:
brum Relocation:
Weigh:
Mix:
Strain:
Transport:
General Cleanliness:
Walls:
Equipmenl
Inventory:
Floors:
Spills:
Engineering Controls:
Ventilation:
Drainage:
Building:
Ventilation:
F-5
-------
Plant I.D. (2 digits) /
Recorder:
TIME IN/OUT OF THE DRUG ROOM
Weigher I.D. f
TOTAL
TIME
HR. MIN
GRAND TOTAL
FOR PAGE
F-6
-------
Plant I.D. (2 digits) /
Recorder:
MASS OF EACH WEIGHING OF POWDERED DYE OR CHEMICAL
DYE OR CHEMICAL WEIGHED
BATCH TICKET NAME
MASS OF WEIGHINGS (SPECIFIC UNITS
Complete this information if this is a continuation sheet page of pages
F-7
-------
Plant I.D. (2 digits) /
Recorder:
Bulk
Sample I.D.
Batch Ticket
Name
• vritxunuix vit,a |fj nnu unc.nj<
FULL TRADE NAME
.HL>&
D/C
(I) WEIGH]
Lot No.
• "^^^^•.^^^•i
:!D |
Sup
^^^^^•^
^^^•^^^
Complete this information if this is a continuation sheet page of pages.
F-8
-------
PEPSQ.-.AL OIR SA.MPL;VC
SHEET
jComp.iny.
i
'*'orkf>r Monitored:
i
'Job Title/Work Duties:
ampljng Performed by:
Plant in No.
D
I Date (Mo/Dn/Vi )
™'o"
Job Title/Work Duties:
fSampUng Performed by: " —
,odel Su.b,
"" - *
SAMPLiN"-, tQflPMEXT AND TALI BRAT ION
"""" «
riowm-tfr Serial Number:
iFJowmeler Cijjbratioa
Calibration Traceable To:
PUMP «2
•Kcrrv*BB
PERSONAL
Mode] No.:
FIELD SAMPLING DATA
Sanple ID Nuaber:
Sample Start Time:
•Sample Stop Ti«e:
Puap flow Rate: (L/M1NJ
jSample Air Volume: (.3)
I
'Signature:
F-9
-------
STATIONARY AREA AIR SA/Tf ING DAT* SHEET
ID No
Sampling Perfor»cd~~iy:
= e = m-tKEBBBT-erc = B» _-*«__.
Flowmcter Model Number:
! Fiowmeter Serial Number""
Date fMo/Pn'vrj
FJowmeter CaJibrat ion Dit-
Calibration Traceable to~
| Sampling Pump Serinl
Sampling Pump Model Numh«--
Pre-sampling
Post-sarapJing FJowrnte:
r ...
• Sample ID Number
Field Blnnk
Sampler Location
Sample Start Tiae:
I Sample Stop Time:
Sample Duration: (mini
Flow Rate: (L/minJ
Sample Air Volume: («3)
I Calculations Checked by:
Date:
Date:
F-10
-------
A.VALVTICAL DATA
Sample ID Number:
' CRAvi^nc"NALYsIs"e"""" "" -'
- = ^.,, .„ n...., mt
FiJter Prcweight
I fU1."- BJanK
FiJter Postweighf
Sample
^ • ---
BJank Correction
Adjusted Weight
SignntureTT
C.ili;uJatioris Checked by
Extraction Dite:
SrrCTR0rHOTOHrTR,C ANALYSIS
* ........ """"1 ........ - ......... -, .....
D.ita File Number
r
SaB|PJe Prepared by:
Total Absorbance:
Corri-ctea Total Absorbnnc«:
Absorptivity:
Oye Estinate: (up)
Corrected Oye Estimate: (,,Kj
Estimate. Airborne Dyes: (ug/>3)
~^_______^_^_
Data Reference Number:
Signatureis):
:«lcuJations Checkel by-
FJJter Spike Reentry.-
F-ll
-------
0 D 0»i
(Sn lock * tap far Ex»lonorlo* a/ Nut*,*)
CHAIN Of CUSTODY OR TUCU.IUTY tfCORD
Ue«tio«___
Contain* No..
Dot* »( Pic'd Sampling .
TyiM»/S«Mpl«
Kao,uir»«i
S«ol Intact
Dm
(MHali)
Oar.
Tim*
F-12
-------
2.
3.
4.
5.
6.
7.
8.
10.
11.
12.
13.
Check chain of custody form.
Traceability log.
Enter project and task number.
the^first and last samples were collected that are recorded
Enter sampling location: plant nane and/or city.
'-. etc.
Enter storage r.auiren-ents, i.e.. «t ice, dry ice. in plastic bags,
Enter entire sample number.
Enter any other sample description required.
Enter other sample identification, i.e., Tenax tube numbers
Enter name or initials of person collecting sample.
checkmark in the appropriate box I f
level should be confined at '
be noted and dated under the
Dieted, enter the a.V5
«
inventoried, place a
HgVid' the "
s in the level should
at the beainning of each day? Ch«k Sff Jj^8Sples must be inspected
record the inspection date in 17 fSd IM^Ii fhe,SMl is intict in «•
inspection, the seal may be broken -" 18' A£ter se«1
"
a
to the laboratory cufitoSTi ?« e MinPles until
laboratory cu.todi.n SiS 22i2 £f JhJ'to™" Jransfer' P«"id. the
the crew chief. copies or the forms and give the originals to
F-13
-------
APPENDIX G
DATA TABLES FROM IN-PLANT QUESTIONNAIRE
G-l
-------
DATA TABLES TO BE PROVIDED IN NEXT DRAFT
G-2
-------
APPENDIX H
INDIVIDUAL SITE LABORATORY ANALYSES
H-l
-------
ANALYSES TO BE PROVIDED IN NEXT DRAFT
H-2
-------
APPENDIX I
SENSITIVITY ANALYSIS OF DYE CONCENTRATION
ESTIMATES DUE TO MEASUREMENT ERROR
1-1
-------
1.0 INTRODUCTION
This section is a sensitivity analysis of the effect of
measurement error for airborne dye concentration. In this
analysis ANOVA techniques were applied to estimate the component
of total sample variance due to measurement errors and to the
underlying population variance. The estimated population variance
is then applied to generate percentiles of the population
distribution which are not biased upward by the additional
variance due to measurement errors.
While many ANOVA techniques have been proposed, the standard
one-way, random effects model will be considered here^1^. The
one-way model is applied to separate the total sample variance
into two components: one for the variation of exposure levels
within-plants; and another for variations across (or between)
plants. In the former case, it is appropriate to consider the
deviations of each observation from the mean of all observations
for that plant. In the latter case, we consider the squared
deviation of the plant means from the overall mean for all plants.
Measurement errors are shown to affect both the within-
plant and across-plant components of variance in the survey.
Within-plant variance is affected by the traditional
field/laboratory airborne chemical measurement sources of
variation- such as variations in flow rates of collection devices,
filter efficiency, efficiency of sample recovery and extraction,
and errors inherent in laboratory measurements using spectrometric
devices. These errors add to the spatial variation of dye
concentration within the weighing area and the variation due to
left- and right-handed workers. The total of such within-plant
variations is collectively estimated in the within-plant component
of variance.
1-2
-------
In the survey, there is also an effect of measurement error
in the estimated across-plant variance. This second type of
measurement error relates to the problem of measuring the total
concentration of multiple dyes with a single spectrometric
measurement. To accomplish this, it is necessary to know the
relative amounts of each dye in the collected sample. In the
survey, these relative amounts are based on physical measurement
of the amount of each dye weighed. Due to variations in
"dustiness" of each dye weighed, the true airborne relative
concentrations may differ from the weighed concentrations.
This type of measurement error affects all measurements
within a plant in a similar fashion, and is not a component of the
within-plant variation. This type of measurement error is
contained in the estimated across-plant component of variance.
Hence, the total variance across plants is larger than the time
population variance across plants.
The purpose of ANOVA is to identify the within- and across-
plants components of variance. Additional simulations^2^ were
necessary to estimate the amount of measurement error affecting
the across-plants component of variance. A final estimate of the
underlying population variance is obtained by subtracting the
variance due to the across-plant type of measurement error from
the total across-plant variance.
In the following section, the ANOVA procedures for
estimating total, within, and across-plant variances are
summarized. In the final section of this report, the across-
plant component of variance is decomposed into one component which
represents across-plant measurement error, and the remaining
component is identified as the underlying population variance.
Population statistics are then presented based on the estimated
population mean and variance.
1-3
-------
2.0 ANOVA RESULTS
The final survey data set contains measurements of airborne
concentrations tabulated both on an active dye basis and on a
commercial dye basis. The analysis of variance was conducted
separately for each set of measurements. While the general
discussion is phrased in terms of active dye concentrations,
analogous results hold for the commercial dye concentration
measurements. The active dye measurements contain one set of two
observations (left and right) in each of 22 plants, yielding a
total of 44 data points. Sampling weights are available for
estimating population parameters for all plants and separately for
all workers.
Frequency plots of the unweighted observations on an active
dye basis are shown in Figures la and Ib. Figure la depicts the
original data, while the frequency plot of the natural logarithm
of the observations are shown in Figure Ib. In both figures, left
and right measurements are shown separately. Figures 2a and 2b
show equivalent plots on a commercial dye basis. Examination of
the figures demonstrates that the distribution of the logarithm of
the observations is approximately a normal distribution. The
figures also demonstrate that there is no apparent difference
between the left and right measurements in terms of mean or
variance.
Based on the above observations, the following random
effects log-linear model is adopted:
1) Natural numbers - multiplicative model
Yi,j = M Ai Ei,'j (i = 1,..., NI;J = left, right)
where the *i»j denote the original airborne concentration
measurements in plant i, on side j, using the active dye basis.
1-4
-------
FIGURE la
M
Ln
i
o
Ul
12
11
10
8
7
9
4
3
2
1
0
ACTIVE DYE
A
/!;!
/?
/!
<.03
<.25
<.45
<.53
<.73 >.73
COHCEHTFAT10M
LEFT CSS* RICMT
-------
FIGURE Ib
6
3
Id
2- m
1 -
LOG (ACTIVE DYE)
<-3.3 <-3 <-2.5 <-2 <-
_
CT7I LEFT
LOC(CONCENTHXnON)
»WCKT
-------
FIGURE 2a
COMMERCIAL DYE
H
.4 <.43 <.3 <.33
<.73 >.73
•• CONCENTRATION
1771 LEFT ESS WCHT
-------
FIGURE 2b
00
LOG(COMMERCIAL DYE)
<-3.3 <-3 <-2.3 <-
_ LOC(COHCEKTRATK)H)
CT71 LEFT
>0.3
-------
The fixed parameter M will be discussed below. The terms Ai
denote random (lognormal) multiplicative effects for variation
across plants, and the Ei,j denote random (lognormal) variations
within plants. Taking the logarithms of equation (1) yields:
2) Logarithms - additive model
yi,j = m + ai + ei»j
where yifj = log (Yi,j),
m = log (M),
ai = log (Ai), and
ei,j = log (Eifj).
In Equation (2), the a^ are normally distributed random
effects across plants with mean zero and variance sa2. The ei,j
are normally distributed error terms within plants with mean zero
and variance s2. The fixed parameter m represents the overall
mean of the logarithms of the data. By the theory of the
lognormal distribution, the fixed parameter M (=em) in Equation
(1) above represents the overall median of the observations
expressed in original numerical form. The parameter M is also an
estimate of the median of the underlying population distribution,
because the ai and ei,j have mean zero. The symbol sa2 denotes
the across-plant variance component, while the symbol s2 denotes
the within-plant component of variance. As noted in the
introduction above, sa2 includes both the variance due to the
population distribution and due to the across-plant component of
measurement error. The purpose of the ANOVA presented in this
section is to estimate s2 and sa2. In the following section, our
simulation-based estimate of the across-plant measurement error is
removed from sa2 to yield an unbiased estimate of the population
variance.
Note that all estimates produced by the ANOVA procedure are
statistics calculated from the logarithms of the observations.
1-9
-------
Use of logarithms implies that sa2 estimates the across-plant
variance in the logarithms and s2 estimates the within-plant
variance in the logarithms. Use of logarithms in ANOVA is
equivalent to analyzing the percent variations in the original
observations. Thus, the variance of the logarithms translates to
the mean square percent variation in the original numbers, and the
mean of the logarithm translates to the median of the original
numbers. The mean of the original numbers is a function of both
the median and the variance estimates, due to the skewness of the
lognormal distribution. All results presented in this section are
in terms of the logarithms. Similarly, the across-plant component
of measurement error estimated in the following section is
presented in terms of logarithms. The final results for
population characteristics will be presented in original numerical
form, however.
The log-linear random effects ANOVA model in equation (2)
above was estimated using both weighted and unweighted logarithms.
The use of sampling weights detracts from the simplicity of the
required calculations for unweighted ANOVA. First, we discuss the
unweighted analysis procedure, and then present these results.
This section concludes with an analysis of the weighted results.
Calculation Procedures for Unweighted ANOVA
A. Within plants:
1. Plant i mean
Yi" = (YiTight + Yi'left)/2
2. Squared deviations
dirj2 = (yi,j - yi,.)2.
3. Mean squared error estimate
1-10
-------
MSEW = SSW dw,
where SSW is the simple average of the squared
deviations within-plants
SSW = _1_ J_2 d2i,jf
2ni i j
and dw is the correction factor for the available
within-plant degrees of freedom,
dw = ninj/ni(nj - 1) = 2
4. Within-plant component of variance
s2 = MSEW
B. Across plants:
1. Overall mean
ni
2. Squared deviations
2
- - y.r- )
3. Mean squared error across plants
MSEa = SSA da,
where SSA is the simple average of the squared
deviations across plants
1-11
-------
SSA = _1_ £ d2if.
and da is the correction factor for the available
across-plants degrees of freedom
da = ninj/(ni - 1) = 44/21 = 2.095
4. Across-plant component of variance
sa2 = (MSEa - s2)/2.
Inspection of the unweighted ANOVA results in Tables la and
Ib confirms the following conclusions:
a. The within-plant variation is small compared to
across-plant variation (3.4 percent on an active dye
basis and 3.7 percent on a commercial dye basis).
b. The variance within plants is approximately equal, in
both a commercial or active dye basis.
c. The variance component across plants is 1.306 on an
active dye basis and 1.202 on a commercial dye basis.
This component includes both the population variance
and the across-plant measurement error.
Calculation Procedures for Weighted ANOVA
The four steps described above for obtaining the across-plant
and within-plant variance components proceed in an analogous
fashion for weighted ANOVA, with the simple averages replaced by
weighted averages, and squared deviations from the mean replaced
with weighted squared deviations from the weighted mean. Because
the weights w^ on both observations within a plant are identical,
the weighted mean within plants equals the unweighted mean within
1-12
-------
plants.
The following steps summarize the calculations performed for
the weighted analysis of variance:
A. Within plants:
1 . Plant i weighted mean
2. Squared deviations
3. Weighted mean square error, within plants
MSEa* = SSA* da,
where SSA is the weighted average of the squared
deviations within plants,
SSA* = Z wi(di*,j)2/njX wi
and da is the same correction factor for the
available within-plant degrees of freedom as in the
unweighted analysis of variance.
4. Within-plant component of variance
s2 = MSEW*
1-13
-------
B. Across plants:
1. Overall weighted mean
2. Squared deviations
* . 2 * *
i ,.r = (yi .. - y .,.
3. Weighted mean squared error, across plants
MSEa = SSA* da,
where SSA* is the weighted average of the squared
deviations across plants
SSA* = J_ wi(di*,.)2/Z wi
and da is the same correction factor for the
available across-plants degrees of freedom as in the
unweighted analysis of variance.
4. Across-plant component of variance
sa2 = (MSEa* - s2)/2
Results for the weighted ANOVA based on establishment
weights on an active and commercial dye basis are presented in
Tables 2a and Table 2b. Similar results using weigher level
weights are presented in Tables 3a and 3b.
1-14
-------
3.0 PRESENTATION OF FINAL RESULTS
The across-plant variance estimated in Section 2.0 above
contains both the population and the across-plant measurement
variances.
Simulations by NRI of the errors induced by possible
variations of the airborne dye mixture from the weighed dye
proportions were used to provide an estimate of the variance due
to measurement error. These calculations are shown in Table 4.
In this table, an estimate of the across-plant measurement
variance is computed from the .05 and .95 percentiles of the
simulation results. A estimate of the across-plant measurement
variance was obtained based on the lognormal model using the
formula
smfi = log [Pi(.95)/Pi(.05)]/2(1.64)
for each simulation in plant i. The resulting estimates of the
standard deviations were then averaged to produce an average
standard deviation
^m = 1 y *»m, j •
The measurement error variance was then obtained as s2m. As
shown in Table 4, these variances are quite small, roughly the
same size as the within-plant variance component obtained in
Section 2.0 above.
To generate the final population statistics, the quantity s2m
was subtracted from the across-plant variance component from the
ANOVA:
2 _ 2 2
s pop ~ s a ~s m '
where all terms are variances of logarithms.
/
1-15
-------
The theory of the lognormal distribution was then applied to
estimate population parameters from the overall mean of the
logarithms and the population variance s2pop. Results of these
calculations are shown in Tables 5a and 5b, for active and
commercial dye, respectively. The difference between adjustment
for across-plant measurement error and no adjustment is high-
lighted by including the estimates obtained directly from the
across-plant variance, sa2. The effects of adjustment for the
across-plant measurement variance is quite small.
After reviewing the compact of measurement error estimates it
was decided not to adjust estimates of airborne dye concentration
because of measurement error. This was done for two reasons.
First, the impact of measurement error was quite small
in this case;
In addition, the adjustment technique described above is
very complicated. It was decided that the extra
complexity introduced by the method was not worth the
increased difficulty in explaining the results
considering that the estimates in general were changed
to two significant digits.
1-16
-------
FOOTNOTES
[1] See, for example, Snedecor and Cochran, Statistical Methods,
6th Ed., Chapter 10.
[2] These simulations, performed by Midwest Research Institute,
are presented in the final section of this report.
1-17
------- |