International Technical Conference Proceedings
June 22-25,1992
Session N—Environmental ill
-------
ATMOSPHERIC RELEASES OF HEXAVALENT CHROMIUM PROM HARD CHROMIUM
PLATING OPERATIONS
By Mitchell S. Hall, John D. Dietz,
C. David Cooper, Roger L. Wayson, and Doug Bauman
University of Central Florida
4000 Central Florida Boulevard
P.O. Box 25000
Orlando, Florida 32816-0450
Introduction
The University of Central Florida Department of Civil and
Environmental Engineering is investigating methods for improved
estimation of chemical releases which require reporting under
provisions of SARA Title III (Toxic Release Inventory, Form R).
This paper describes results from a research project initiated in
the Fall of 1990. The preliminary objective was to help the
surface finishing industry to identify and solve specific
problems with completion of Form R. A problem would generally be
defined as a combination of a process (e.g., hard chrome plating)
and a release (e.g., fugitive air releases of hexavalent
chromium). It was concluded that the majority of environmental
releases (water and sludges) are subject to some form of
monitoring in conjunction with an operating permit. Accordingly,
monitoring data is usually available to suppqrt completion of
Form R. The principal exception to this generalization concerns
air emissions, particularly fugitive emissions.
••*
The initial process selected for field .characterization was
hexavalent chromium releases from hard -fchrome plating. The
sampling effort involved simultaneous quantification of fugitive
emissions and emissions through existing tank ventilation systems
into the stack.
Stack Results - Hexavalent Chromium
Discussion of Data Sources
Chromium may exist in several valence states, each with its own
physical, chemical, and toxicological properties. While Cr(III)
is naturally occurring, is relatively non-toxic (and in fact is
an essential dietary mineral), Cr(VI) is classified by the U.S.
EPA as a known human respiratory carcinogen . One main source of
Cr(VI) is the electroplating industry where ,Cr(VI) mists are
emitted from hard chrome plating baths. Other plating processes
such as trivalent chromium plating and chromic.acid anodizing are
not believed to be significant emitters of toxic pollutants.
Extensive studies have been done, primarily by the EPA , to
evaluate atmospheric releases of chromium and to determine the
efficiency of control devices. Facilities selected for emission
testing were representative of hard chromium electroplating
operations based on the size of the plating tanks, the types of
parts plated, and the plating bath operating parameters.
767
-------
Data from eight of these EPA tests (where all operating
parameters were defined) were reviewed and analyzed to
characterize stack emissions of Cr(VI). Two facilities were
visited by the project team to supplement the EPA data and to
complete field measurement characterization of fugitive
hexavalent chromium emissions. One plant was visited by both the
EPA and our project team.
Stack sampling of Cr(VT) was accomplished using an EPA Modified
Method 13B isokinetic sampling train. Indoor ambient air
sampling for the two sites occurred simultaneously using the
fugitive sampling device described in a subsequent section of
this manuscript.
Modeling Effort
Electroplating operations require an electrical current for
plating deposits. Only about 10 to 20 percent of the current
applied is used to deposit chromium on the item plated. Eighty
to ninety percent is consumed by the evolution of hydrogen gas at
the cathode with the resulting liberation of gas bubbles. When
these gas bubbles rise to the surface, they burst, causing
releases of- chromic acid mist into the air. The rate of gas
formation is a function of the chemical or electrochemical
activity in the tank and increases with the amount of plating
work in the tank, the strength and temperature of the solution,
and the current densities in the plating bath. "
The variability in hexavalent chromium stack emissions documented
in the literature may be attributed to the multi-dimensional
nature of the process. A large number of variables influence the
amount of Cr(VI) released prior to reaching a control device.
Items of information were extracted, if available, for each test
run from the database for the purpose of modeling uncontrolled
stack emissions. These items were:
* Test Site Identification
* Reference Document (if from EPA study)
* Process Information (Type of process, Tank dimensions and
volume, Freeboard, Hexavalent chromium tank
concentration, Bath temperature, Agitation, Cover type
(if used), Mist suppressant (if used), Ampere-hours,
Part type and surface area, and Type of ventilation
system)
* Emissions Information (Exhaust gas volumetric flow rate,
Hexavalent chromium concentration at control device
inlet, Hexavalent chromium mass emission rate at inlet,
Type of control .device, Ambient temperature, and
Duration of test)
7CQ
-------
Uncontrolled Emission Factor.Development
Review of the data from the EPA literature indicated that there
is little relationship between uncontrolled emissions and most
process factors. The following combinations of data provided the
most promising emission factor correlations for uncontrolled
Cr(VI) emissions:
1. milligrams Cr(VI) vs ampere-hours
2. milligrams Cr(VI) vs ampere-hours x bath area
3. milligrams Cr(VI) vs plated area x time
4. milligrams Cr(VI) vs ampere-hours x bath volume x bath
Cr(VI) concentration
5. milligrams Cr(VI) vs ampere-hours x bath volume
Due to qualitative differences in conditions between plants,
development of an overall predictive model using all test data
was not successful. Attempts with multiple variable regression
using the entire data set failed to produce any attractive
correlations.
However, some correlations were more notable. Uncontrolled
emissions appeared to be directly correlated with ampere-hours
when data from single facilities were used. Combination of data
from the eight EPA facilities and the two facilities visited by
our project team suggested a common linear relationship between
Cr(VI) emission rate and ampere-hours. However, this combined
model exhibits considerable scatter.
The emissions rate data was correlated with the product of the
ampere-hours and the plating bath volume (combination number 5
mentioned above). The results are presented in Figure 1. The
observations are segregated .into two groups which a're defined by
the following relations:
Cr = 49,000 + 0.0016 x Amphr x Vol (1)
Cr = 2,900 + 0.00065 x Amphr x Vol (2)
in which Cr = mass emission of hexavalent chromium
prior to control devices (milligrams)
Amphr = power consumption (Ampere-hours)
Vol = plating tank volume (Liters)
Factors to explain the discrimination between the two groups have
not been identified. The more conservative model (Equation 1;
predicting higher emission rates) could be used for general
prediction of hexavalent chromium emissions. An approach
incorporating the entire data set is still being investigated.
769
-------
Figure 1 - Uncontrolled Stack Emissions For Hexavalent Chromium
350
O
1
V.
.Q)
1
•9
*5j
I
to
fro
cc
CO
~
_c
I-
D
50 100 150 200 250 300 350
Ampere-Hours x Bath Volume (Ah-Liters)
(Millions)
400
Fugitive Results - Hexavalent Chromium
SamplingDevice
The fugitive emission samples were collected using a sampling
train based on the EPA EMTIC CTM-OQ6 Method . This method
applies to the determination of Cr(VI) in stack emissions from
decorative and hard chrome plating facilities. The approach used
in the EMTIC Method involves collection of Cr(VI) from a
ventilation stack into a one- quart glass impinger. This
impinger contains 250 ml of- a 0.1 N sodium hydroxide buffer
solution to trap and stabilize the Cr(VI) that is collected. A
second impinger in series is filled with a desiccant to capture
any moisture prior to the air entering the dry gas meter.
The impinger sampling method used in this study differed from
that of the EMTIC Method in five major ways:
(1) The sampling inlet was fixed at 52 inches from the
ground in order to approximate breathing zone concentrations,
rather than attached to a probe which measured stack
concentrations;
770
-------
(2) Two glass impingers with alkaline buffer solution were
used in series, rather than only one, to ensure capture of all
sampled Cr(Vl) mist;
(3) The dry gas meter was placed between the last glass
impinger and the critical orifice to prevent pump oil from
fouling the dry gas meter, rather than last in the sampling
train;
(4) The dry gas meter was no longer exposed to ambient air
pressure so a temperature .gauge and pressure gauge were placed at
the meter inlet and outlet, respectively; and,
(5) No ice bath was used for the impingers since moisture
condensation in the ambient air was not measured.
A schematic of the sampling train is shown in Figure 2. The
components of the train were available commercially but some
fabrication was required. A sturdy wood support housing made of
pine 2 x 4's and 2 1/2 inch wood screws was designed to be
compact and portable to accommodate the needs of the project.
Figure 3 shows a photograph of one of the five sampling trains
constructed at the University of Central Florida Department of
Civil and Environmental Engineering.
Figure 2 - Sampling Train Schematic
Thermometer
Pressure Gees
o
Dry Gcs
NX -s t & r-
Silico Gel
GIG:
I fr\ p i r~i
-------
Figure 3 - Photograph of Sampling Train
Sampling Criteria
Air was drawn through the impingers using 1/4 hp vacuum pumps.
To sample Cr(VI), a constant sampling rate of 0.67 acfm for a 2
to 3 hour period was used. This sampling rate was controlled by
use of a critical orifice downstream of the sample line.
The system was designed to achieve a liquid phase concentration
of Cr(VI) in the scrubber solution of 0.4 to 1.0 mg/L, well in
excess of the 0.016 mg/L detection limit. For the stated
conditions, detection limits of air phase concentrations were
less than 10 ug/m .
Sample Analysis and QA/QC
Kexavalent chromium was analyzed in accordance with Standard
Methods Method 3500-Cr D (17th Edition) . The Cr(VI) ions
captured in the alkaline scrubber solution were complexed with
diphenylcarbazide solution. A colorimetric procedure was
performed in the field within 24 hours of sample collection. The
dry gas meters were calibrated using a wet test meter. The
entire sampling train was tested for leaks before each operation.
Samples were stored and transported in plastic containers which
were thoroughly cleaned in nitric acid and rinsed with distilled
water before use. The analytical program included the following
quality control checks: Duplicates, spikes, field blanks, and
check standards.
772
-------
Field Sampling - Plant 1
The facility at this site processed large industrial rolls.
Plating activity occurred in only two of three chromic acid
tanks, but ambient air samplers were placed near all tanks
because they were still heated and ventilated. One operating
tank was covered by a steel plate except for top openings through
which cylinders to be plated could be inserted. Forced
ventilation was used on three sides of the tank, but the tank
cover allowed air to enter only the airspace above the liquid
through the top openings mentioned. Ventilation of all tanks
occurred by use of a duct system which was routed to a single
outlet stack. A sampler was placed both upstream and downstream
of 'expected air currents across this tank. One sampler was
placed away from all tanks to allow measurement of background
concentrations. The other two tanks were not covered, even
during plating operations, and one sampler was placed at the
"downwind" side of each of these tanks.
FieldSampling - Plant 2
The facility at the second site processed aircraft engine parts.
The shop had five plating tanks near the west wall of a large
room. The room was sectioned off for ventilation purposes.
During normal operations, all plating tanks-were covered with a
plastic sheet in an effort to stop fugitive air emissions. The
tanks were ventilated by a duct system which was routed to a
single outlet stack. Four samplers were placed next to four of
the most active plating tanks and a fifth sampler was placed east
of the tanks for background sampling. Ambient air sampling was
done at one tank with and without the plastic cover in place.
Room Air FlowRates
Determination of mass emission rates (stack and fugitive) were
made using the measured ambient air concentrations and measured
air flow rates. Air velocity measurements were completed for
doorways and windows which represented inlet air sources to the
plating room at each of the two facilities we studied. These
measurements showed that a negative pressure in the room resulted
in a net air flow entering the room through all windows and
doors. All air flow leaving the room was through the stack and
ventilation system. For these circumstances, fugitive emissions
would eventually be re-captured into the stack or deposited onto
surfaces (floors, walls) in the plating room.
Air flow rates in the plating facilities were determined with
three separate techniques:
1. Estimation based on engineering specification of the
ventilation system, not including the stack which was
sampled;
2. Velocity measurements in the sampled stack; and,
3. Velocity measurements at windows and doors.
773
-------
Table 1 below shows results of the airflow measurements made at
the two plants visited by our project team.
Table 1 - Air Flow Rates at Studied Plants
Plant
1
2
Air Flow Rates (scfm)
Inputs
Windows/Doors
14500
26500
Outputs
Stack
6100
13460
Ventilation
9000
13350
Total
15100
26810
As shown in Table 1, it can be concluded that the total air
exchange in the rooms is essentially equal to the total
ventilation flow rate. Thus, escape of any Cr(VI) mists into
the room air will either fall out in the room or will eventually
be carried out through the stack or room ventilation system.
CrfVI) Sampling Results
Four sampling runs were performed at each site; two on each of
the two days of sampling that occurred at each location. • Based
on an assumed production schedule of 2000 hours per year for both
plating facilities, a final summary of the average results of
stack and fugitive sampling is provided in Table 2 below.
Table 2 - Fugitive and Uncontrolled Stack Cr(VI) Results
Plant
1
2
Cr(VI) Cone
(ug/std cu m)a
Stackb
10040
414
Fugitive
< 4
< 6
Annual Emissions
(Ib Cr (VI) /year)
Stackb
460
41.7
Fugitive
< 0.5
< I
amicrograms per standard cubic meter.
bStack emissions are prior to reaching the control
device.
Fugitive emissions were more difficult to quantify than stack
emissions for the following reasons.
1. Fugitive sample results were generally below detection
limits. This led to the use of a conservative approach in
calculating ambient Cr(VI) concentrations. The subsequent
analysis is predicated on air concentrations at the detection
limit of 4 ug/std cu m in Plant 1 and 6 ug/std cu m in Plant 2.
774
-------
It is believed that actual concentrations were much less than
this limit, thus the resulting calculations would yield a decided
overestimate of the actual result.
2. Fugitive emissions are collected along with a portion of
room air via the facility ventilation system and may be carried
outside through other parts of the plant via stacks which have no
control devices. However, some of the Cr(VI) mists in the room
air are captured by the plating room stack ventilation system and
are incorporated into stack emissions. The calculations of
fugitive emissions shown in Table 2 were based on an extremely
conservative approach in which the total room ventilation was
assumed to contribute to the fugitive emission rate.
For the stated conditions, the annual emission rates due to
fugitives from Plants 1 and 2 are < 0.5 Ib/year and < 1 Ib/year,
respectively. The values for both facilities are believed to
greatly exceed actual fugitive releases for the reasons cited
above. Even so, for the purposes of completion of Form R, this
estimate of the fugitive releases would be reported in the
minimum category (zero to one pound per year).
Fugitive emissions from Plant 1 were negligible in comparison to
stack mass flow rates. In fact, for a typical scrubber
efficiency of 99%, stack exit emissions of Cr(VI) would be 4.6
Ib/year for plant 1, which is still an order of magnitude greater
than the corresponding fugitive emissions. The only fugitive air
sample which was above detection limits was obtained immediately
downstream from an uncovered process tank which experienced loss
of vapors over the unventilated end of the tank.
The results from Plant 2 are in general agreement with the
findings from the first facility; fugitive releases do not appear
to represent a significant portion of the total air emissions.
Only one ambient sampler showed a measurable Cr(VI)
concentration. This result was obtained near an operating tank
during testing with the plastic covering removed. The
inexpensive plastic sheets covering the tanks appeared to
minimize emissions as suggested by the employees.
Cgncl_us_iQ.ns
While the' EPA has done extensive work in the last decade to
quantify controlled and uncontrolled stack emissions of
hexavalent chromium from hard chrome plating facilities, no
documented efforts were made to measure fugitive emissions. This
study represents an assessment of airborne fugitive hexavalent
chromium concentrations at these facilities. In an effort to
develop a model for stack emissions of Cr(vT) , EPA data were
reviewed and a correlation for chromium emissions was reported
versus ampere-hours and plating bath volume.
775
-------
A modification of a stack sampling train was constructed to
accommodate fugitive sampling efforts made at two hard chrome
plating facilities. The levels found at both site, for the most
part, were below detection limits of 4 ug/std cu m and 6 ug/std
cu m for the 1st and 2nd plant, respectively. The annual
emission rates from Plants 1 and 2 are < 0.5 and < 1 Ib Cr(VI)
per year, respectively. For the purposes of completion of Form
R, these estimates of the fugitive releases would be reported in
the minimum category (zero to one pound per year).
Acknowledgements
The information in this article has been funded wholly or in part
by the U.S. Environmental Protection Agency under Cooperative
Agreement CR-817586-Q1-0 to the AESF. Financial support from the
AESF is also gratefully acknowledged.. This article has not been
subjected to the Agency peer review process and does not
constitute an endorsement by or necessarily reflect the view of
the Agency.
References
1. U. S. Environmental Protection Agency, "Health Assessment
Document for Chromium: Final Report," EPA-600/8-83-
014F, Environmental Criteria and Assessment Office,
Research Triangle Park, NC, 1984.
2. U. S. Environmental Protection Agency, "Chromium Emissions
from Chromium Electroplating and Chromic Acid Anodizing
Operations-Background Information for Proposed Standards,"
Volume I, NESHAP Preliminary Draft, Office of Air Quality
and Standards, Research Triangle Park, NC, Dec. 1990.
3. U. S. Environmental Protection Agency, "Draft Modified
Method 13B - Determination of Hexavalent Chromium
Emissions from Decorative and Hard Chrome Electroplating,"
Environmental Monitoring Systems Laboratory, Office of
Research and Development, Cincinnati, OH, Dec. 1990.
4. U. S. Environmental Protection Agency, "Determination of
Chromium Emissions from Chromium Electroplaters," EMTIC-
006 Conditional Test Method, Prepared by Frank Clay,
Emission Measurement Branch Technical Support Division,
OAQPS, EPA, 8 Nov. 1990.
5. American Public Health Association, "Standard Methods for
the Examination of Water and Wastewater," 17th Edition,
Washington, DC, 1991.
776
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completi
1. REPORT NO.
EPA/600/A-93/069
2.
3.
4. TITLE AND SUBTITLE
5. REPORT DATE
ATMOSPHERIC RELEASES OF HEXAVALENT CHROMII
HARD CHROMIUM PLATING OPERATIONS .
FROM
6, PERFORMING ORGANIZATION CODE
M> s> Hallj 0> D> DietZ5 c> David cooper,
R, L. Wayson, and D. Bauman
8. PERFORMING ORGANIZATION REPORT NO,
9, PERFORMING ORGANIZATION NAME AND ADDRESS
University of Central Florida
4000 Central Florida Boulevard
P.O. Box 25000
Orlando, FL 32816-0450
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR-817586-01-0
12. SPONSORING AGENCY NAME AND ADDRESS
Risk Reduction Engineering Laboratory—Cincinnati, OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Presentation & Proceedings
14, SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer — Mark J. Stutsman, (513) 569-7776
Presented at American Electroplaters &Surface Finishers Society's Annual SUR/FIN Conf,
held in Atlanta, GA on June 22-25, 1992. Proceedings pp. 767-776.
16. ABSTRACT
The University of Central Florida Department of Civil and Environmental
Engineering is investigating methods for improved estimation of chemical releases
which require reporting under provisions of SARA Title III (Toxic Release Inventory,
Form R).
This paper describes results from a research project initiated in the Fall of
1990. The preliminary objective was to help the surface finishing industry to
identify and solve specific problems with completion of Form R. A problem would
generally be defined as a combination of a process (e.g., hard chrome plating) and a
release (e.g., fugitive air releases of hexavalent chromium). It was concluded that
the majority of environmental releases (water and sludges) are subject to some form
of monitoring in conjunction with an operating permit. Accordingly, monitoring data
is usually available to support completion of Form R. The principal exception to
this generalization concerns air emissions, particularly fugitive emissions.
The initial process selected for field characterization was hexavalent chromium
releases from hard chrome plating. The sampling effort involved simultaneous
quantification of fugitive emissions and emissions through existing tank ventilation
systems into the stack.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Group
Toxicity
Chrome
Plating
Toxic Release Inventory
Hexavalent Chromium
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report/
UNCLASSIFIED
>1. NO. OF PAGES
12
20. SECURITY CLASS (This page I
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R»v. 4-77) PREVIOUS EDITION is OBSOLETE
------- |