oEPA
United States j::::jj •:::::: :::::: ':::jjj:::::::::
Environmental Protection :::::: ::::: ::::" ::::"::::!!
Agency Sh Emission Redjction
Partnership for the Magnesium Industry
Characterization of Emissions and
Occupational Exposure Associated with
Five Cover Gas Technologies for
Magnesium Die Casting
Office of Air and Radiation
Office of Atmospheric Programs, Climate Change Division
EPA 430-R-07-008
-------
Characterization of Emissions and Occupational Exposure
Associated with Five Cover Gas Technologies for Magnesium Die
Casting
Scott Bartos
United States Environmental Protection Agency
Climate Protection Partnerships Division
1200 Pennsylvania Avenue, NW (6202J)
Washington, DC 20460
August 2007
EPA 430-R-07-008
-------
Table of Contents
Page
Acknowledgement vii
Executive Summary ES-1
1.0 Introduction 1-1
2.0 Principles of FTIR Monitoring 2-1
2.1 The Spectrum Analysis Method 2-1
2.2 Creating the Spectrum Analysis Method 2-1
2.3 Reference Generation 2-4
3.0 The Extractive FTIR and Other Sampling Systems 3-1
3.1 Hydrogen/Oxygen Analyzer 3-5
3.2 Quadrapole Mass Spectrometer (QMS) Analyzer 3-5
3.3 FTIR Calibrations and System Checks 3-6
3.3.1 Cell Leak Checks 3-7
3.3.2 Infrared Detector Linearity Checks 3-7
3.3.3 Noise Equivalent Absorbance or Signal-to-Noise Ratio Tests 3-7
3.3.4 Path Length 3-7
3.3.5 Spectrometer Frequency and Resolution Checks 3-7
3.3.6 Spectral Background 3-8
3.3.7 Sample Cell Exchange Rate 3-8
3.3.8 FTIR Measurement Error 3-8
4.0 Test Results 4-1
4.1 MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas 4-2
4.2 AM-Cover™ using HFC-134a with CDA Carrier Gas 4-5
4.3 SF6 with CDA Carrier Gas 4-8
4.4 SO2 with CDA Carrier Gas 4-11
4.5 Frozen CO2 4-13
4.6 Determination of Dilution 4-16
4.7 Occupational Exposure Monitoring 4-19
4.7.1 Observed Compounds During Occupational Exposure Monitoring 4-21
in
-------
Table of Contents (continued)
Page
5.0 Conclusions 5-1
5.1 Cover Gas Test Observations 5-1
5.1.1 MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas 5-1
5.1.2 AM-Cover™ using HFC-134a with CDA Carrier Gas 5-2
5.1.3 SF6 with CDA Carrier Gas 5-2
5.1.4 SO2 with CDA Carrier Gas 5-2
5.1.5 Frozen CO2 5-3
5.2 Cover Gas Destruction 5-3
5.3 Global Climate Change Potential Discussion 5-5
5.4 Uncertainty Discussion 5-7
IV
-------
List of Tables
Page
ES-1 Cover Gas Average Concentrations and Observed Destruction ES-3
ES-2 Observed Compounds From Occupational Exposure Ambient Air Monitoring ES-6
1-1 Test Schedule for FTIR Sampling at the Lunt Facility 1-3
1-2 Magnesium Die-casting Machine Parameters 1-4
2-1 Analysis Method Parameters for Major Contaminants and Spectroscopic Interferants 2-3
4-1 Data Summary for MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas 4-3
4-2 MDL Summary for all Monitored Compounds During MTG-Shield™ Tests 4-4
4-3 Data Summary for AM-Cover™ using HFC-134a with CDA Carrier Gas 4-6
4-4 MDL Summary for all Monitored Compounds During AM-Cover™ Tests 4-7
4-5 Data Summary for SF6 with CDA Carrier Gas 4-9
4-6 MDL Summary for all Monitored Compounds During SF6 Tests 4-10
4-7 Data Summary for SO2 with CDA Carrier Gas 4-12
4-8 MDL Summary for all Monitored Compounds During SO2 Tests 4-13
4-9 Data Summary for Frozen CO2 4-15
4-10 MDL Summary for all Monitored Compounds During Frozen CO2 Tests 4-16
4-11 Cover Gas Compound Occupational Exposure Details 4-20
4-12 Monitored Compounds from Occupational Exposure Ambient Monitoring 4-22
5-1 Percent Destruction for Cover Gas Testing 5-4
5-2 Comparison of 100-year GWP Estimates from the Intergovernmental Panel on
Climate Change (IPCC) Second (1996) Assessment Report 5-5
5-3 Normalized GWP Comparison of Measured Emissions from Inside the Crucible
Headspace 5-8
5-4 GWP (Weighted by Flow Rate) Comparison of Measured Emissions from Inside the
Crucible Headspace 5-9
-------
List of Figures
Page
1-1 Die Casting Machine Crucible at Lunt Manufacturing 1-5
2-1 Reference Generation Hardware Configuration 2-5
3-1 Sampling System Schematic 3-2
3-2 Sample Locations for the Tested Crucible 3-4
3-3 Nova 340WP Oxygen and Hydrogen Analyzer 3-5
3-4 The QMS used to Monitor N2 Inside the Crucible for Dilution Calculation 3-6
4-1 QMS Response Curve for 0-50% Nitrogen 4-17
4-2 QMS Response to N2 During MTG-Shield™ Sampling 4-18
VI
-------
Acknowledgement
The analytical measurement and research effort to prepare this report was funded by the
U.S. Environmental Protection Agency under contract GS-10F-0124J to ICF International. The
authors wish to express their appreciation and thanks to Lunt Manufacturing and their staff,
specifically Helmut Brandt and John Kinart, for contributing not only their facilities, but their
valuable assistance and advice, to this measurement study. The support of Advanced
Magnesium Technologies (AMT), Matheson Tri-Gas/Taiyo Nippon Sanso Corporation (TNSC),
Polycontrols Technologies, Inc. and 3M™ for providing their cover gases, expertise and trial
staff for this study is also gratefully acknowledged.
vn
-------
Executive Summary
This measurement study was conducted to evaluate the greenhouse gas (GHG) emissions
and occupational exposure associated with five cover gas technologies used in a cold chambered
die casting operation. Sulfur hexafluoride (SF6) is widely used for the protection of molten
magnesium, but with the goal of eliminating the use of SF6 in this application by 2010, the
magnesium industry and U.S. Environmental Protection Agency (EPA) have been evaluating the
use of alternative gases. This study expands upon previous research by evaluating additional
alternative cover gases and including an occupational exposure component.l This study
examined the use of AM-Cover™ using HFC-134a (supplied by Advanced Magnesium
Technologies), MTG-Shield™ using Novec™ 612 (supplied by Matheson Tri-Gas, Taiyo Nippon
Sanso, and 3M™), sulfur dioxide (802) (gas blender supplied by Polycontrols Inc.), and frozen
carbon dioxide (CO2) and SFe (both provided by Lunt Manufacturing) on a single cold
chambered magnesium die casting machine located at a Lunt Manufacturing facility in
Hampshire, Illinois. Each cover gas mixture was evaluated under identical process and machine
operating parameters. With the exception of frozen CC>2, each cover gas was injected into and
extracted from the crucible headspace under similar parameters to characterize emissions and
byproducts as the cover gases interact with the melt surface and undergo thermo-degradation.
The results reported were from measurements taken inside the crucible headspace and from
worker exposure/ambient air sampling points. Table ES-1 summarizes some of the details and
results from the crucible head space component of the study. Measurements were conducted
using multiple cover gas mixtures, and in the case of frozen CC>2, a different injection location
and physical phase of chemical. The cover gas destruction rates listed in Table ES-1 have been
corrected for crucible dilution effects.2 Table ES-2 summarizes the results from continuous
monitoring of ambient air at two worker stations associated with the die casting machine.
MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas
The primary destruction byproducts measured while running MTG-Shield™ using
Novec™ 612 with CO2 as a cover gas were CO, COF2, C3F8, C2F6, CHF3 and HF (see Table 2-1
for a listing of chemical formulas and compound names). The C$6 concentrations ranged from
below detectable limits (BDL) to 13 parts per million by volume (ppmv) and the COF2
concentrations ranged from BDL to 36 ppmv, depending on injected concentration and stability
of the crucible headspace. HF concentrations were detected at levels of 7 to 450 ppmv for the
range of mixtures evaluated. Higher feed gas concentrations resulted in higher concentrations of
HF. Additional compounds detected included CH4, C2H4, and CH2O. CH4 was detected above
1 Characterization of Cover Gas Emissions from U.S. Magnesium Industry Die Casting Operation, March 2004.
2 The term destruction is utilized throughout the remainder of this report to represent the thermo-degradation and
disassociation of the cover gas agent resulting in byproduct formation and melt protection.
ES-1
-------
ambient concentrations and C2H4 and CH2O spiked during ingot loading. Low levels of N2O (<
5 ppmv) and NO2 (0.4 to 5 ppmv) were detected during casting and tended to increase during
ingot loading periods.
AM-Cover™ using HFC-134a with CDA Carrier Gas
The primary destruction products measured while running AM-Cover™ with HFC-134a
with a compressed dry air (CDA) carrier gas were CO, HF, and COF2. HF concentrations were
measured from 448 to 1,199 ppmv with higher values correlating to higher concentrations of
F£FC-134a in the feed gas. Concentrations of COF2 ranged from 16 to 59 ppmv. The time series
concentration plots for these and other compounds are presented in Appendix A. They illustrate
that additional destruction products, such as CH2O and C2H4; are formed with the addition of
ambient air during the ingot loads. Some compounds, such as CH2O, NO, N2O and NO2 also had
background levels inside the headspace that sharply increased during ingot loading.
ES-2
-------
Table ES-1. Cover Gas Average Concentrations and Observed Destruction
Cover Gas
Mixture
Components
Novec™612/C02
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/C02
Novec™612/C02
HFC-134a/CDA
HFC-134a/CDA
SF6/CDA
SF6/CDA
SO2/CDA
S02/CDA
S02/CDA
S02/CDA
S02/CDA
SO2/CDA
SO2/CDA
Frozen CO2
Date
8/22/06
8/22/06
8/22/06
8/22/06
8/23/06
8/23/06
8/23/06
8/24/06
8/25/06
8/24/06
8/24/06
8/28/06
8/28/06
8/28/06
8/28/06
8/28/06
8/29/06
8/29/06
8/29/06
Time
0910-1005
1008-1135
1138-1301
1347-1425
0906-0958
1108-1248
1252-1556
1510-1736
0810-1215
0939-1311
1806-1902
1105-1130
1133-1218
1222-1348
1351-1500
1504-1552
0845-1001
1005-1155
1651-1829
Cover Gas Mixture
Flow3
(1pm)
-36
-36
-36
-36
-36
-36
-36
^0
-40
-35
-35
-39
-39
-39
-39
-39
-39
-39
Liquid CO2 at 100 psie
Cover
Gas
Delivery
Conc.a
(ppmv)
800
600
400
300
200C
200d
150
4,200
3,600
3,000
3,000
10,000
8,500
7,000
6,000
5,000
5,000
4,000
1,000,000
Cover Gas
Measured
Cone.
(ppmv)
198.8
142.5
55.4
3.2
75.8
64.2
30.8
1,198.3
810.7
1,966.1
1,932.2
6,086.6
5,521.3
4,652.5
4,157.3
3,401.3
3,042.0
2,518.2
957,390.1
Cover Gas
Destruction1"
74%
76%
86%
99%
61%
67%
79%
71%
77%
32%
34%
37%
33%
32%
29%
30%
37%
35%
4%
aAs provided by Lunt Manufacturing, AMT, Polycontrols, and Matheson Tri-Gas/TNSC
b Dilution corrected values from Table 5-1
0 Data collected prior to a dross
d Data collected after a dross
e Liquid CO2 dewar at 100 psi
The detection of CH2O and C2H4 was sporadic, with a few measurable spikes occurring
during ingot loading. Detection of SF6 was also observed at low concentrations and is a residual
from previously used SF6 cover gas.
SF6 with CDA Carrier Gas
The primary destruction byproducts measured while using SF6 with CDA as a cover gas
were FTP and SO2. FTP concentrations ranged from 1 to 49 ppmv for the first and second tests,
respectively. N2O and NO2 levels were also vastly different between the two test periods. CH2O
levels were somewhat consistent between tests with the second test slightly less than the first.
Low levels of C2H2 were also observed during the first test. CH4 was observed at an average of
approximately 3 ppmv during both tests, slightly above the ~2 ppmv normally seen in ambient
air.
The primary difference between the 2 test episodes is the time the SF6 had been purging
the crucible head space prior to sampling. The SF6 mixture had been flowing for approximately
ES-2
-------
17 hours before the first test, whereas the second test had started immediately after a cover gas
changeover to SFe. The longer purge period prior to testing had a pronounced effect on the
emissions, as the HF, N2O, NC>2 and CO concentrations were all significantly lower. These
compounds also follow a pattern of concentration increases during ingot loadings. The
temperature instability of the crucible headspace gas (due to dilution with cold ambient air
during ingot loadings) and/or the displacement of cover gas by excess ambient air may have
played a role in this effect.
SO2 with CDA Carrier Gas
There were few destruction byproducts that could be attributed to SC>2. Most of the
observed compounds were either a carry-over effect of residual chemicals from the previous
cover gas used in the crucible (e.g., SFe and FTP), ambient air components (F^O, CC>2, CH/t),
destruction byproducts formed from ambient air dilution during ingot loading (CH2O and C2H4)
or nitrogen oxides formed from the CDA carrier gas. FI^SC^ was not measured at concentrations
above its minimum detection limit (0.051 ppmv) within the crucible headspace.
Frozen CO2
The final cover gas tested was frozen CC>2. As the gas cooled through expansion, it froze,
forming solid phase CC>2 which was gravity fed into the crucible. The obvious difference
between the frozen CC>2 and all the other cover gases was the method of injection. Compressed
CC>2 was delivered to the nozzle at a pressure of 100 psi from a liquid CC>2 dewar. Due to the
inconsistencies associated with delivery and crucible headspace pressures, effects of dilution also
differed. However, the quadrupole mass spectrometer (QMS) also monitored N2 concentrations
during the CC>2 cover gas testing to enable a calculation of dilution (described in Section 4-6).
The delivery/formation of the frozen CC>2 was controlled and optimized by regulating the
injection into the crucible with a solenoid valve.
CC>2 had very few decomposition by-products, with the exception of CO. A very small
amount of C2H4 was consistently present throughout the testing. SO2, HF and SF6 were observed
during the testing, but were present as a carry-over effect of residual chemicals from previous
cover gases used in the crucible. The presence of low concentrations of nitrogen oxides was
most likely due to thermal decomposition of ambient air entering the crucible through leaks and
ingot loading.
ES-4
-------
Observed Percent Destruction for Cover Gases
Table ES-1 lists the destruction estimates for all cover gases examined. The destruction
estimates, which are corrected for dilution effects (i.e., the effects of air ingression into the
crucible headspace), are calculated as the percent difference between the delivery concentration
and the measured concentration in the crucible headspace. Average destruction estimates for
Novec™ 612, HFC-134a, and SO2 were on the order of 77%, 74%, and 33%, respectively.
Frozen CC>2 had the lowest observed destruction at 4%. In comparison, destruction estimates for
SF6 were on the order of 33% for this study.
The destruction rates estimated for SF6 in this study were significantly higher than what
was estimated during previous research (on the order of 10%). This is likely due to the much
lower feed gas SF6 concentrations utilized in this study, and reduced levels of dilution resulting
in destruction having a larger share of the reduction in measured concentration.
Occupational Exposure Monitoring
Since each cover gas used in this study can result in emissions that may be harmful to
exposed workers, monitoring of the ambient air near worker breathing zones was performed. A
second FTIR was used to monitor the breathing air at two locations based on their probability for
worker activity. This component of the study was especially relevant to SC>2 due to its stringent
occupational exposure limits. For Novec™ 612, HFC-134a, and SF6-based cover gases, the
monitored zone was at the end of the casting process where the part was robotically dropped to
an area where the worker inspects the part and places it on a pallet for transfer to the finishing
process. This station was occupied by a worker about 50% of the time and was approximately
18 feet from the crucible. For SO2 and frozen CO2 cover gases, the area above the crucible was
monitored near the ingot loading door. Although the worker activity is lighter at this location,
the potential for elevated exposure concentrations is the highest due to direct contact with
escaping crucible gases. Table ES-2 lists the monitoring results for the primary compounds of
concern that were observed along with their established permissible exposure limit (PEL) and
short term exposure limit (STEL). SF6 was consistently present throughout the monitoring due
to its usage at the other casting operations occurring throughout the facility. With the exception
of a single measurement when SC>2 was being tested, continuous monitoring of these worker
areas found concentrations of the primary compounds of concern that were either BDL or well
below PEL and STEL values.3
1 All applicable safety precautions (e.g., operational procedures) should be followed when using SO2.
ES-5
-------
Table ES-2. Observed Compounds from Occupational Exposure Ambient Air Monitoring
Cover Gas
Date
Zone**
Permissible Exposure Limit (PEL)
Short-term Exposure Limit (STEL)
MTG-Shield™
8/22
1
Max
Average
Novec™612
(ppmv)
150
n/a
BDL
BDL
HFC-134a
(ppmv)
1,000
n/a
n/a
n/a
SF6
(ppmv)
1,000
n/a
0.08
0.06
SO2
(ppmv)
2
5
n/a
n/a
HF
(ppmv)
3
6
BDL
BDL
CO
(ppmv)
50
400
11.60
7.07
COF2
(ppmv)
2
5
BDL
BDL
CH2O
(ppmv)
0.75
2
BDL
BDL
MTG-Shield™
8/23
1
Max
Average
BDL
BDL
n/a
n/a
0.19
0.06
n/a
n/a
BDL
BDL
14.45
6.80
BDL
BDL
BDL
BDL
AM-Cover™
8/24
1
Max
Average
n/a
n/a
0.050
0.020
0.16
0.10
n/a
n/a
BDL
BDL
15.33
5.28
BDL
BDL
BDL
BDL
AM-Cover™
8/25
1
Max
Average
n/a
n/a
0.078
0.038
0.29
0.18
n/a
n/a
BDL
BDL
6.48
2.75
BDL
BDL
BDL
BDL
SF6
8/24
1
Max
Average
n/a
n/a
n/a
n/a
0.28
0.18
BDL
BDL
BDL
BDL
9.48
2.75
BDL
BDL
BDL
BDL
SO2
8/28
2
Max
Average
n/a
n/a
n/a
n/a
0.23
0.19
1.60*
0.14
BDL
BDL
7.22
1.21
BDL
BDL
BDL
BDL
S02
8/29
2
Max
Average
n/a
n/a
n/a
n/a
0.07
0.03
BDL
BDL
BDL
BDL
11.49
7.06
BDL
BDL
BDL
BDL
Frozen CO2
8/29
2
Max
Average
n/a
n/a
n/a
n/a
0.23
0.16
n/a
n/a
BDL
BDL
11.46
4.55
BDL
BDL
BDL
BDL
BDL = below detectable limits
n/a = not applicable
*Occurred during an instance when the ingot loading door was open for a prolonged period due to an ingot loading malfunction.
"Zone 1 was located at the process end where the part is robotically dropped to the worker.
Zone 2 was located near the ingot loading area of the crucible.
ES-6
-------
Potential Climate Impact
A key factor in evaluating alternative cover gas compounds was their composite global
warming potentials (GWPs) as compared to SF6. Global warming potentials are based on the
heat-absorbing ability and decay rate of each gas relative to that of carbon dioxide. The GWP
provides a construct for converting emissions of various gases into a common measure,
denominated in carbon dioxide equivalents. For each cover gas compound and its applicable
destruction byproducts (e.g., CsFg, C2F6), a composite global warming impact estimate was
developed using the IPCC second assessment report GWP values.4 The overall GWP-weighted
gas emission rate for each cover gas regime was estimated using the measured average
concentrations of each gas, their molecular weights and the delivery cover gas flow rates. This
resulted in a normalized CC>2 emission equivalent for each alternative cover gas that could be
directly compared to the CC>2 emission equivalent of SFe.
Based on this approach, results indicate that the MTG-Shield™ using Novec™ 612, AM-
Cover™ using HFC-134a, and frozen CC>2 have a GHG emission impact that is at least 98%
lower than SF6. The 862 cover gas has an associated GHG emission impact that is 99% lower
than SF6.
4 IPCC, Climate Change 2001: The Scientific Basis. Intergovernmental Panel on Climate Change, 2001, Cambridge
University Press. Cambridge, U.K.
ES-7
-------
1.0 Introduction
This report presents and interprets the results of a series of cover gas emissions
measurements on a single cold chambered magnesium die casting machine. Measurements were
conducted by URS Corporation (URS) at a Lunt Manufacturing facility located in Hampshire,
Illinois between the 22nd and 29th of August, 2006. Measurements were made in a continuous
and real-time fashion with an extractive-type Fourier Transform Infrared (FTIR) spectroscopic
system, an extractive Quadrapole Mass Spectrometer (QMS), and an extractive-type oxygen (O2)
continuous emission monitor (CEM).
The focus of the study was to assess destruction byproducts and emissions for five
different cover gases on a die casting machine operating under identical operational parameters.
Cover gases are used to prevent surface oxidation and burning of the molten metal during
processing. The five cover gases evaluated in this study were: 1) MTG-Shield™ using Novec™
612 (supplied by Matheson Tri-Gas, Taiyo Nippon Sanso, and 3M™), 2) AM-Cover™ using
HFC-134a (supplied by Advanced Magnesium Technologies), 3) SO2 (gas mixer supplied by
Polycontrols Technologies, Inc.), and 4) frozen CO2 and 5) SF6 (supplied by Lunt
Manufacturing). Objectives for this study were:
• To characterize the greenhouse gas emissions (GHG) from various cover gas regimes
used for magnesium melt protection during die-casting operations. Emissions
measurements by FTIR were employed to identify the gaseous fluorides, acids and
perfluorocarbons (PFCs) that may result from cover gas decomposition.
• To identify any detectable occupational exposure emissions associated with the use of
each cover gas. Worker exposure areas were monitored for known and suspected
compounds and reaction by-products using a long path FTIR.
• To determine the extent of destruction for each cover gas. As a control measure, each
cover gas was injected into the confines of a single process crucible during identical
casting operations.
• To determine the amount of dilution from ambient air into the crucible during normal
operation. Direct measurement (of O2 by CEM) and analysis of nitrogen (by QMS)
during the Novec™ 612 studies provided an accurate determination of head space
dilution, to be applied to all other cover gases (of nearly equal injection flow rates),
except CO2 (of unknown injection flow rate).
• To determine the GHG emissions from the cover gas technologies and overall
reduction in GHG emissions attributable to the use of MTG-Shield™ using Novec™
612, AM-Cover™ using HFC-134a, SO2 and CO2 as compared to SF6.
1-1
-------
The measurement schedule, sampling locations, and test conditions are summarized in
Table 1-1. The die casting process parameters are summarized in Table 1-2. The measurements
were conducted under these conditions during identical casting activity for all cover gases
employed. An HPM 400 die-casting machine (#1), fabricating an automotive lock body housing,
was chosen for the testing. Figure 1-1 depicts the die casting crucible that was tested at the Lunt
Manufacturing facility.
Testing was carried out using independent cover gas flow controls through the existing
gas distribution apparatus affixed to the crucible cover. The MTG-Shield™ cover gas was
generated via a dedicated supply cabinet in which cylinders of liquid Novec™-612 are heated to
supply a constant concentration gas stream to a gas mixing panel using a gas blending cabinet
built and operated by Matheson Tri-Gas/TNSC. The AM-Cover™ cover gas was generated
using a gas blender provided by Lunt Manufacturing. The 862 cover gas was generated using a
gas blending cabinet built and operated by Polycontrols. SFe was supplied using the existing
centralized gas distribution system at the facility. There were five cover gas injection points
evenly spaced in a ring around the top of the crucible. Each injection point was modified by
Lunt Manufacturing engineers by installing a 45 degree nozzle that directed the flow of cover
gas onto the melt surface. There was also a sixth cover gas injection point at the top of the ingot
loading hatch. The directional nozzle modification to the existing cover gas delivery system
resulted in improved melt protection effectiveness for four of the cover gases examined. Frozen
CC>2 was delivered through a system where it was gravity fed to the melt surface using a single
injection point near the center of the crucible lid.
1-2
-------
Table 1-1. Test Schedule for FTIR Sampling at the Lunt Facility
Date
8/22/06
8/22/06
8/22/06
8/22/06
8/22/06
8/23/06
8/23/06
8/23/06
8/23/06
8/24/06
8/24/06
8/24/06
8/24/06
8/25/06
8/25/06
8/28/06
8/28/06
8/28/06
8/28/06
8/28/06
8/28/06
8/29/06
8/29/06
8/29/06
Time
0740
0910-1005
1008-1135
1138-1301
1347-1425
0906-0958
1000-1100
1108-1248
1252-1556
0939-1311
1312
1510-1736
1806-1902
0730
0810-1215
1015
1105-1130
1133-1218
1222-1348
1351-1500
1504-1552
0845-1001
1005-1155
1651-1829
Cover Gas Mixture
Components
Cover Gas Mixture Flow3
(1pm)
Cover Gas
Delivery
Cone. a
(ppmv)
Ingot Type
Dross
Novec™-612/C02
Novec™-612/CO2
Novec™-612/CO2
Novec™-612/CO2
Novec™-612/CO2
-36
-36
-36
-36
-36
800
600
400
300
200
AZ91D
AZ91D
AZ91D
AZ91D
AZ91D
Dross
Novec™-612/C02
Novec™-612/C02
SF6/CDA
-36
-36
-35
200
150
3,000
AZ91D
AZ91D
AZ91D
Dross
HFC 134a/CDA
SF6/CDA
-40
-35
4,200
3,000
AZ91D
AZ91D
Dross
HFC 134a/CDA
-40
3,600
AZ91D
Dross
S02/CDA
S02/CDA
SO2/CDA
SO2/CDA
SO2/CDA
SO2/CDA
S02/CDA
Frozen CO2
-39
-39
-39
-39
-39
-39
-39
Liquid CO2 at 100 psi
10,000
8,500
7,000
6,000
5,000
5,000
4,000
1,000,000
AZ91D
AZ91D
AZ91D
AZ91D
AZ91D
AZ91D
AZ91D
AZ91D
aAs provided by Lunt Manufacturing, AMI, Polycontrols, and Matheson Tri-Gas/TNSC.
1-3
-------
Table 1-2. Magnesium Die Casting Machine Parameters
Parameter
Facility
Furnace Temperature (°F)
Ingot Weight (Ibs)
Furnace Capacity (Ibs)
Alloy Type
Mg Casting Rate (seconds/part)
Mg Pump Type
Mg Shot Weight (Ibs)
Metal Throughput (Ibs/hr)
Product
Molten surface area (sq ft)
Ingot Loading
Machine #la
Lunt Manufacturing: Hampshire, IL
1,260
23
600
AZ91D
33
Cold Chambered Gas Displacement
0.8
85.6
Automotive Lock Body Housing
12.6
Automatic Feed
aAs provided by Lunt Manufacturing
1-4
-------
Figure 1-1. Die Casting Machine Crucible at Lunt Manufacturing
1-5
-------
2.0 Principles of FTIR Monitoring
Almost every chemical compound absorbs infrared (IR) light to some degree in a
particular region of the mid-infrared spectrum. These absorption properties can be used to
identify and quantify chemical compounds in a complex mixture of gases. As stated by Beer's
Law, the magnitude of a compound's IR absorbance is directly proportional to the product of its
concentration in the mixture and the sample cell optical path length. This product is otherwise
known as the compound's optical depth. The extractive FTIR instruments used by URS are able
to achieve parts-per-billion (ppb) detection levels because the optical path length within the
measurement cell is magnified many times by reflecting the IR beam between a series of mirrors
before it reaches the detector. The mirrors provide a fixed optical path length best suited to the
gas mixture being sampled. In this case, optical path lengths of 20.1 meters (for worker
exposure monitoring) and 5.1 meters (for crucible head space monitoring) were utilized.
2.1 The Spectrum Analysis Method
An infrared spectrum analysis is performed by matching the features of an observed
spectrum to those of reference gases of known concentration. If more than one feature is present
in the same region, then a linear combination of references is used to match the compound
feature. The standards are scaled to match the observed band intensities in the sample. This
scaling also matches the unknown concentrations. An infrared spectrum can be collected and
analyzed in approximately one second, but spectra are normally averaged over one- to five-
minute integration periods to produce adequate signal-to-noise limits and ppb detection levels.
The scaled references are added together to produce a composite that represents the best
match with the sample. A classical least squares mathematical function is used to match the
standards' absorption profiles with those of the observed spectrum in specified spectral analysis
regions. The compounds of interest together with compounds expected to cause spectral
interference are included in the analysis region.
2.2 Creating the Spectrum Analysis Method
The spectrum analysis methods used for the tests at the Lunt Manufacturing facility were
developed by URS scientists by selecting the spectral regions and sub-regions that are least
affected by primary IR absorbers (EtzO and CC>2, in this case) while also producing the best
detection limit possible for the target compounds. Target compounds are initially determined
prior to sampling and are based on previous tests and the cover gas composition. However,
many destruction byproduct gases were found during data analysis requiring many iterations of
data processing and interpretation. Typically, an analysis method is iteratively refined by using
it to analyze a representative set of infrared spectra while varying the method. The optimum
method is indicated when both the 95% confidence levels and the bias on the individual
2-1
-------
compounds are minimized. Table 2-1 lists the range of references included in the analysis
method used by the FTIR systems for the tests. Each reference is described in terms of its
optical depth (i.e., concentration times cell path length (ppmv-meters) range.
After setting up the FTIR instruments on-site, signal-to-noise ratio (SNR) assessments
were performed. This was determined by measuring the noise equivalent absorbance (NEA) of
each FTIR system while sampling nitrogen. The NEA is derived by ratioing two consecutive
single beam spectra to produce a "zero" spectrum, then measuring the peak-to-peak absorbance
at a frequency region of interest. This represents the noise level of the instruments under field
conditions. By determining the concentration level for each contaminant that scales down its
analyzed spectral features to the NEA (representing a SNR of 1 or better), the compound's SNR-
limited minimum detection limit (MDL) can be estimated.
Due to the complexity of the sample matrices, the detection limits reported in Section 4
for each compound were calculated using one of two different methods. The first method was
used when spectroscopic interferences were taken into account for those contaminants that have
overlapping absorption features and an increase in their noise based MDLs was expected. To
determine this MDL, an actual data set of spectra where the target analyte was not observed (via
spectral validation by a URS spectroscopist), but all other process gases were present, was used.
Within this data set, any positive concentration observed for the target analyte would be a
mathematical anomaly created by interferences. Three times the standard deviation of the data
set is an approximation of the method limited MDL that contains over 99% of all expected data
points within the noise scatter. This method is preferred to the theoretical "noise based"
detection limit since it accounts for the effects of interferences. The calculation is a more
conservative and practical calculation and therefore was used wherever possible. The second
method was conducted when the analyte of interest was always present in the sample stream or
there were insufficient data points for the previously mentioned method.
A theoretical noise-based detection limit was determined by comparing the peak-to peak
noise value calculated from the NEA spectrum to the absorbance intensities and optical depth of
the lowest concentration reference used in the analytical method. The equation below shows the
calculation for theoretical noise-based detection limits. Note that this MDL is multiplied by two
as a conservative estimate (to include at least 95% of all noise scatter). For all MDL
calculations, a peak-to-peak noise of 1 x 10"3 absorbance units was used.
Peak - to - Peak Noise Reference Absorbance
2 x MDL x Path Length Reference Pathlength x Reference Concentration
2-2
-------
Table 2-1. Analysis Method Parameters for
Major Contaminants and Spectroscopic Interferants
Chemical
Formula
H20
C02
SF6
C2H2F4
C3F7C(0)C2F5
SO2
CO
HF
COF2
C2H2
C2H4
C2F6
CF4
CHF3
CH3F
CH4
OF2
CH20
CH2O2
NO
N20
NO2
H2SO4
S03
Compound
Water
Carbon Dioxide
Sulfur Hexafluoride
HFC-134a
Novec™ 612
Sulfur Dioxide
Carbon Monoxide
Hydrofluoric Acid
Carbonyl Fluoride
Acetylene
Ethylene
Hexafluoroethane
Carbon Tetrafluoride
Trifluoromethane
Methyl Fluoride
Methane
Oxygen Difluoride
Formaldehyde
Formic Acid
Nitric Oxide
Nitrous Oxide
Nitrogen Dioxide
Sulfuric Acid
Sulfur Trioxide
SF6
(ppmv-
meters)
2.89-22.3*
70-2,110
58-92,701
n/a
n/a
518-10,415
26-20,358
1-2,000
50-5,000
111-5,550
86-2,576
448-1,119
5.6-1,120
112-560
177-1,182
87-21,119
1,750-14,000
92-1,838
76
53-2,043
102-1,019
34-1,544
n/a
n/a
HFC-134a
(ppmv-
meters)
2.89-22.3*
2.06-33.45*
56-280
9,700-27,575
n/a
n/a
1,500-11,200
1-2,000
50-5,000
111-5,550
86-2,576
448-1,119
5.6-1,120
112-560
177-1,182
87-21,119
1,750-14,000
92-1,838
76
53-2,043
102-1,019
34-1,543
n/a
n/a
Novec™ 6 12
(ppmv-meters)
2.89-22.3*
137-510*
56-280
n/a
99-991
n/a
784-20,358
1-2,000
50-5,000
111-5,550
86-2,576
448-1,119
5.6-1,120
112-560
177-1,182
87-21,119
1,750-14,000
92-1,838
76
53-2,043
102-1,019
34-1,543
n/a
n/a
SO2
(ppmv-meters)
2.89-22.3*
70-2,110
56-280
n/a
n/a
518-35,770
26-3,863
1-2,000
n/a
n/a
86-2,576
n/a
n/a
n/a
n/a
87-21,119
n/a
92-1,838
76
53-2,043
102-1,019
34-1,543
164
1400
CO2
(ppmv-meters)
2.89-22.3*
137-510*
56-280
n/a
n/a
518-35,770
784-20,358
1-2,000
50-5,000
111-5,550
86-2,576
n/a
n/a
n/a
n/a
87-21,119
n/a
92-1,838
76
53-2,043
102-1,019
34-1,543
n/a
n/a
*Expresses in %-meters since high concentration references were required.
In some instances compounds absorb infrared light in regions that were interfered by
higher concentration compound absorbances, providing large differences in MDL from one data
set to another. An example would be CF4. The MDL for CF4 in the HFC-134a data set was 3.8
ppmv but only 0.013 ppb in the SF6 data set. CF4 was difficult to detect in a sample stream
containing high HFC-134a concentrations since its strongest absorbance peak is in a region that
HFC-134a absorbs infrared in. Therefore, an alternative region was chosen where the
absorbance intensity of CF4 was much weaker, thus increasing its detection limit. In the SFe data
set the strongest CF4 absorbance peak was in a relatively clean region, resulting in the lower
MDL.
2-3
-------
2.3 Reference Generation
Since the use of HFC-134a and Novec™ 612 within the magnesium industry is relatively
new, FTIR references were required to be generated for both gases. Additionally, high
concentration CC>2 references were required since concentrations greater than 85% CC>2 were
observed when used as a carrier gas for Novec™ 612, and during the frozen CC>2 injections. The
Novec™ 612 references were obtained from a previous study5. These references were generated
from certified gas standards, made gravimetrically on NIST certified scales, from pure Novec™
612. The standard was certified at ± 2 percent at 201 ppmv Novec™ 612 by HP Gas Products
located in Baytown, Texas. The standard was diluted at 5 different levels with gaseous nitrogen
and FTIR reference samples were measured via the following procedure:
1. Evacuate and fill the FTIR sample cell with nitrogen 5 times
2. Evacuate the cell using a roughing vacuum pump
3. Add ultra high purity (UHP) nitrogen to a cell pressure of 400-650 torr
4. Add Novec™ 612 gas standard while recording pressure differential
5. Fill cell to 750 torr with N2
6. Measure gas reference
With this approach the pressure was monitored at each reference step with a calibrated
Baratron pressure sensor made by MKS. By knowing the amount of standard added as a
function of pressure, the concentration was calculated by the following equation.
_. . _, . Pressure of the Standard Added _, , _
Reference Concentration = x Bottle Concentration
Total Pressure
HFC-134a references were also made in similar fashion. Each gas reference sample was
saved and used to generate calibration curves that were then applied to the Novec™ 612 and
HFC-134a data. For CC>2, additional gas standards at high concentrations were generated from a
gas cylinder containing 62.92% CC>2 during the previous study3. Using this CC>2 gas
concentration in a longer path length cell (20.1m as opposed to 5.11m), references were
generated such that the CC>2 calibration curve had an effective upper limit of 100% CC>2. Figure
2-1 is a schematic of the configuration used for generating these references.
1 Characterization of Cover Gas Emissions From US Magnesium Die Casting Operations. March 2004
2-4
-------
Pressure
Measurement
MKS
Baratron
Pressure
Measurement
Vacuum
Pump
Gas
Standard
UHP
N2
Figure 2-1. Reference Generation Hardware Configuration
2-5
-------
3.0 The Extractive FTIR and Other Sampling Systems
Two extractive-type FTIR systems were used for measurements conducted at Lunt
Manufacturing. MKS (On-Line) FTIR spectrometers and sample cells were used. The crucible
head space monitoring system included two monel sample probes (3/8" OD), a heated PFA-
grade Teflon extraction line, the On-Line FTIR spectrometer interfaced to a heated, nickel-
coated sample cell, a venturi sample pump, and an exhaust tube. Given this configuration, real-
time monitoring consisted of pulling a continuous gas stream, in equal portions from each
sample probe, through the sampling system into the heated FTIR sample cell. Sample flow was
maintained at approximately 3.2 liters per minute (1pm) by a venturi pump connected to the
outlet of the FTIR cell. A schematic is shown in Figure 3-1. The worker exposure/ambient air
monitoring system consisted of a length of unheated Vi" PFA grade Teflon extraction line, the
MKS On-Line FTIR spectrometer, a heated (35°C), long path length (20.1 m), nickel-coated
sample cell, a venturi sample pump, and an exhaust line. The sample gas was pulled from the
worker breathing zone at approximately 3 1pm through the extraction line and into the sample
cell.
Inside each FTIR cell, a set of optically matched gold-plated mirrors reflects an infrared
beam through the sample gas multiple times. As the beam passes through the sample, the
molecules in the sample absorb some of its energy. After exiting the cell, the infrared beam is
directed to a liquid-nitrogen cooled mercury/cadmium/telluride (MCT) detector, a
photoconductive device that produces an electrical voltage proportional to the amount of infrared
light that strikes it. The strength of the absorption at particular frequencies is a measure of the
compound's concentration. The total distance traveled by the infrared beam inside the cell is the
cell path length, and is an important variable used in determining sample concentrations. For
this project, cell path lengths were fixed at 20.1 m for the worker exposure system and 5.1m for
the crucible head space system.
The crucible FTIR sample cell and extraction lines were maintained at a temperature of
150°C (to prevent any condensation losses and preclude the formation of HF mists). The worker
exposure/ambient air monitoring extractive line was at room temperature and the cell was
maintained at 35°C. For both FTIR systems, cell pressures were continuously recorded during
measurement periods using a pressure sensor calibrated over the 0-900 torr range. Instrumental
resolutions were set to 0.5 cm"1 and signal averaging was performed over 3- and 5-minute
periods for the crucible spectrometer and ambient spectrometer, respectively.
5-1
-------
Exhaust
Mixing "T"
FTIRwith
Heated (150° C)
Cell
Heated (150° C)
Extraction Line
Crucible Furnace
Unheated Monel Sample Prob
Unheated Monel Sample Probe
Figure 3-1. Sampling System Schematic
3-2
-------
A single die-casting crucible was tested for all cover gases and is illustrated in Figure 3-2
along with the sample extraction locations on the crucible lid. These two sample points were
used simultaneously for each set of tests. The probe lengths and connecting tubing were
identical in length and fed into a "T" where both gas streams mixed. The sample ports were
existing openings in the crucible lid normally used for thermocouples, pump assemblies, etc.,
and allowed enough room for the sample probes to be placed through the lid without altering its
design. Simultaneously sampling from two points allowed for a more representative sample of
the headspace area (as compared to sampling from one point at a time). Ideally, a larger
manifold and more sample points would be used, but this was not feasible due to the excessive
crucible lid modifications required.
The two sampling points were equally spaced on each side of the ingot feed door.
Samples at both points were simultaneously extracted through monel tubes inserted into the
headspace. The locations had no bearing on the cover gas distribution regime and, when
combined, provided representative gas mixture from the headspace of the crucible. Also of note
is that the sampling regime, in terms of elevation above the melt surface, was consistent between
the two sample extraction locations. The magnesium ingots were automatically fed to the
crucible as needed, approximately every 15 minutes, so the loading door would open then
immediately close after an ingot was dropped. The door would remain open for less than 10
seconds during this process.
In order to measure the potential occupational exposure concerns associated with the use
of these cover gases, the second FTIR was set up to sample from the two primary worker stations
at this die casting machine. The first location (Zone 1) is near the part drop point for the robotic
arm. At this stage in the die casting process the part is briefly examined by the worker and
placed in a container for transport to the next stage in the production process. A sample line was
run from the FTIR to a steel support beam adjacent to the drop table where the worker generally
stands. The sample point was located 18 feet from the edge of the crucible and 63 inches above
the floor (to simulate a typical respiration height). Samples were taken on a continuous basis
from this point during testing of AM-Cover™, MTG-Shield™, and SF6. The second location
(Zone 2) was 30 inches away and 12 inches above the ingot loading hatch on the crucible lid.
Ventilation around the casting machine (and throughout the whole facility) was
substantial. The machine is approximately 60 feet from a large overhead door opened to the
outside; there were multiple industrial floor fans in the area, and there is a sub-floor fresh air
supply vent located 30 feet from the machine that was directing a continuous flow of fresh air
towards the machine. Additionally, the entire facility has high-volume exhaust fans in the
ceiling that contribute to the area around the machine (and the whole facility) having a very high
air exchange rate.
5-3
-------
Crucible Furnace
Ingot
Loading
Door
Figure 3-2. Sample Locations for the Tested Crucible
3-4
-------
Figure 3-3. Nova 340WP Oxygen and Hydrogen Analyzer
3.1 Hydrogen/Oxygen Analyzer
A Nova Model 340WP
portable analyzer was used over a
2-hour period during a MTG-
Shield™ using MTG-Shield™
cover gas test for the continuous
measurement of oxygen
concentrations. The instrument is
shown in Figure 3-3. This
instrument uses an electrochemical
sensor to measure oxygen over a
range of 0 to ambient levels
(20.9%). A chemical reaction
occurs when the sensor is exposed
to oxygen, resulting in a millivolt
output proportional to the oxygen concentration in the sample gas. This small voltage was used
to display the measured oxygen concentration on the instrument's front panel meter as well as
fed to a portable strip chart. A slipstream was taken from the inlet to the FTIR sampling the
crucible headspace and fed into the 62 analyzer. A continuous reading was recorded onto a strip
chart. Since the sampling system was a closed system and the Novec™ 612 was diluted with
CC>2, the 62 present was assumed to be due to mostly ambient air, but presumably a fraction was
formed/destroyed during reactions between the cover gas mixture and the melt surface, thereby
biasing the dilution results. This behavior warranted the need for quadrapole mass spectrometry
(QMS) to monitor an inert gas, in this case N2, in order to accurately gauge ambient air dilution
effects.
3.2 Quadrapole Mass Spectrometer (QMS)
Quadrapole mass spectrometers (QMS) are often used for residual gas analysis in various
industrial applications. It was incorporated in this study to determine the overall combined
dilution or ambient air entering the crucible, from leaks and ingot loading. Traditionally, mass
spectrometers have been used as research instruments for analysis under vacuum applications.
The recent advancements of the technology coupled with the development of atmospheric
samplers and closed ion sources, has enabled atmospheric sampling via QMS. The "high
pressure" QMS, otherwise know as residual gas analyzers (RGAs) are smaller, more robust, and
much more portable than their laboratory predecessors. The RGA used for this testing was a
Leybold Inficon Transpector Closed Ion Source 2 and is shown in Figure 3-4. The Leybold
RGA utilizes a differentially pumped atmosphere-to-analyzer chamber sample interface.
$-5
-------
Figure 3-4. The QMS used to Monitor N2 Inside the Crucible for Dilution Calculation
This allows sample gas, at or near atmospheric pressure, to "leak into" the QMS ionizer
region under controlled rates via precision orifices. All the sample components can then be
simultaneously ionized and mass selected for quantitative analysis. However, the analyzer's
response factors, over all the mass units, are highly dependent on the ionizer's chamber pressure,
energy, etc., so a careful calibration must be performed on-site and under a consistent sampling
configuration. This is presented in Section 4.6.
3.3 FTIR Calibrations and System Checks
A series of on-site calibration and system checks was performed on each FTIR and
sampling system prior to testing to ensure data of known quality. These checks consisted of the
following:
-------
3.3.1 Cell Leak Checks
This test checks the integrity of each cell by pulling a vacuum on it and then monitoring
the leak rate. The acceptance criteria for this test is a leak rate < 2 torr/minute. The evacuated
pressure on each FTIR sample cell did not change over a 1-minute period.
3.3.2 Infrared Detector Linearity Checks
For best results, it must be assured that the infrared detector yields a linear response
throughout a reasonable absorbance range at all the frequencies in a set of test spectra. A
software linearizer is used to continuously adjust the MCT detector preamp signal in order to
achieve the desired response. To optimize the linearizer, background spectra are acquired with
and without a polyethylene film in the IR beam. Comparison of the strongly absorbing
polyethylene bands in the low, mid and high frequency regions against a clean background
enables the processor to appropriately set the linearizer terms (offset, linear, quad, cubic and
delay). This procedure was run prior to the start of testing, and subsequent spectra were visually
checked on a periodic basis to confirm that linearity was maintained.
3.3.3 Noise Equivalent Absorbance (NEA) or Signal-to-Noise Ratio (SNR) Tests
This provides a measure of the system noise, that is the sensitivity of the instrument for
the specified spectral resolution (0.5 cm"1, in this case) and number of scans (256, or two minutes
of signal averaging, in most cases). An NEA test was run upon set-up. The results for both
systems, which were used to assess the field detection limits, were as follows:
20.1m Path Length System 2010
Range = 1000-1100cm'1, RMS Noise=0.22 milliAU
Range = 2450-2550cnT1, RMS Noise=0.25 milliAU
Range = 4200-4300cm'1, RMS Noise=0.72 milliAU
5.11m Path Length System 2030
Range = 1000-1100cm'1, RMS Noise=0.15 milliAU
Range = 2450-2550cnT1, RMS Noise=0.16 milliAU
Range = 4200-4300cm'1, RMS Noise=0.76 milliAU
3.3.4 Path Length
The sample cell used for these tests was geometrically fixed at 20.1 meters for the FTIR
system used for the occupational exposure sampling, and 5.1 meters for the FTIR system used
for sampling the headspace of the crucible.
3.3.5 Spectrometer Frequency and Resolution Checks
A real-time check of frequency position and resolution was performed prior to and
directly after each round of testing by monitoring a specific water absorption line (present in
ambient air). The position of this line must not deviate more than + 0.005 cm"1 from the
3-7
-------
reference value over the course of each test. Likewise, the line width (directly related to
-i
instrument resolution) of this line must not deviate more than + 0.05 cm" from the reference
value over the course of each test.
3.3.6 Spectral Background
A spectral background is essentially a "blank spectrum" in that it does not contain any of
the target compounds present in the sample. It was created by purging the cell with ultra-high-
purity (UHP) nitrogen while collecting a spectrum. This spectrum was then used by the
analytical software to ratio against each sample spectrum to produce an absorbance spectrum for
quantitative analysis. A new spectral background was generated each day prior to the beginning
of testing.
3.3.7 Sample Cell Exchange Rate
Given sampling flow rates on the order of 3 liters/min through either cell during the
testing, a complete sample exchange takes place every 12 seconds for the 5.1 meter cell, and 32
seconds for the 20.1 meter cell. Since spectral signal averaging was conducted over 3- and 5-
minute intervals (for the headspace and worker zone respectively), each collected spectrum
represented an integrated average over multiple sample cell exchanges.
3.3.8 FTIR Measurement Error
As with all analytical devices, extractive FTIR measurements are known to have a given
error associated with them. Steps were taken throughout the measurement process to minimize
sampling error. Sampling error is dependant on many factors including spectral interferences
contained in the sample stream, instrumental noise of the FTIR systems, infrared detector
nonlinearities, and optical depth of references that were applied. Errors were minimized by
applying a series of references at various optical depths to account for any nonlinearities or
dynamic concentration swings in the sample matrix. Spectra were also manually inspected for
qualitative and quantitative validation. As a result of these efforts, the measurements taken in
this study have a level of uncertainty that was significantly lower than ±30 %.
-------
4.0 Test Results
This section presents all the cover gas test data and is broken into sections according to
cover gas type as they were chronologically tested: MTG-Shield™ using Novec™ 612, AM-
Cover™ using HFC-134a, SF6, SO2, and frozen CO2. Additionally, the QMS calibration/dilution
data and occupational exposure testing are included in sections 4.6 and 4.7, respectively. Table
1-1 of Section 1.0 shows the test schedule, cover gas injection flow rates and concentrations used
during testing. Data collected during sampling downtime (i.e., probe taken out while switching
cover gases) or abnormal operation (dressing) were excluded from the data tables, plots and
calculations of this report. For compounds not observed above their respective MDLs a "BDL"
(Below Detection Limit) was reported. There were some data sets that would bracket the MDL,
i.e. had values above and below the detection limits. To calculate average values over these data
sets, a value of half the MDL was used in place of the BDL values. Normally a range for
average values is reported using zero for the low range and the detection limit for the high range.
However, the following data tables are simplified by using the "average value" (half the MDL)
of the detection limit to calculate concentration averages. When calculating the standard
deviation for data sets containing concentrations above and below the MDLs, the raw data was
used in place of the BDL so as not to bias the deviation. Since the standard deviations were
calculated with the ingot loading times included (known dilution from ambient air when the
loading door was opened), they were skewed high and representative of a non-stable system.
Detection limits were calculated as three times the standard deviation while the analytes were not
present in the sample streams (see Section 2 for a detailed explanation). When an analyte was
present, a method based detection limit was determined by using the noise-based equation
defined in Section 2.2.
On several instances immediately after connecting the extraction line to the sample
probes, low concentrations of hydrochloric acid (HC1) were detected that would slowly decay
throughout each testing period. This behavior was independent of cover gas process activity (for
no chlorine was employed in any gas mixtures) and therefore is attributed to chlorine being
stripped from the Teflon core in the extraction line, or possible contamination from another
source.
The data in this section are summarized in Tables 4-1 through 4-12. Appendix A
presents plots detailing data trends and process activities, such as ingot loading, for all the
compounds that were observed above detection limits. BDL data entries were plotted as zero on
the charts in Appendix A. The data summaries in Tables 4-1 through 4-12 represent normal
crucible operation specific to this machine (HWP 400 die casting Machine 1) fabricating a
specific part (automotive lock body housing). The periods when the ingot door was open for
4-1
-------
ingot loading were included in the results presented data tables.6 Comparisons of this data to
other die-casting machines would not be entirely appropriate, since the operational parameters
would most likely be different (thereby affecting the average dilution). However, all the data in
this study was collected on a single crucible/die-casting machine, fabricating the same part,
providing a direct comparison of the five different cover gas technologies.
4.1 MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas
A series of tests was run that sequentially lowered the injected concentrations of Novec™
612 in the MTG-Shield™ cover gas mixture to optimize its performance. The MTG-Shield™
cover gas was used over a two-day period at six different levels for a step-wise approach to
optimization beginning with the highest level (800 ppmv) on the morning August 22nd and
ending with the lowest level (150 ppmv) on the afternoon of August 23rd. Table 4-1 presents the
data summary for the observed compounds at each level tested during normal casting conditions.
Appendix A contains the respective time series plots for the observed compounds. Additional
compounds were monitored but not observed above their MDLs. Table 4-2 presents the MDLs
for all the compounds monitored during the MTG-Shield™ testing. Testing during day 1
indicates that the Novec™ 612 destruction increases as its inlet concentration level decreases.
So, as the inlet concentration decreases, the destruction rate approaches 100% - the optimal
condition for melt protection. During day 2 this trend was repeatable, with the absolute
destruction rates slightly lower.
In addition to CC>2 and unreacted Novec™ 612, destruction byproducts of CO, COF2, HF,
CH4, C2Fe, CHFs, and CsFg were observed. HF concentrations were detected at levels of 7 to
450 ppmv for the range of mixtures evaluated.7 Higher feed gas concentrations resulted in
higher concentrations of HF. Formaldehyde, C2H/t, N2O and NC>2 were also observed, and were
believed to be present mostly as a function of ambient air dilution. For CH2O and C2H4, this was
supported by the observation of slight concentration increases during ingot loading, then sharp
decays to baseline levels or BDL immediately after ingot drops. This effect was less pronounced
for N2O and NC>2. An interesting trend also occurred with CO as the Novec™ 612 was
sequentially decreased. Average CO concentrations were very stable (just over 800 ppmv)
during the 800, 600, and 400 ppmv Novec™ 612 cover gas tests.
6 The crucible headspace was not purged with the new cover gas prior to measurements so the initial readings for
each cover gas include a small protection overlap with the previous one used. The impact of this is likely to be
negligible in terms of overall cover gas performance from a GHG emission perspective.
7 The addition of dry air to the cover gas mixture to minimize unwanted byproducts (and reduce GHG emissions)
was not feasible given the short duration of this study. A more accurately optimized system would likely include a
dry air component.
4-2
-------
Table 4-1. Data Summary for MTG-Shield™ using Novec-612™with CO2 Carrier Gas
Min (SOOppmv)
Max (SOOppmv)
Average (SOOppmv)
Stdev
H20
(%)
0.2
0.3
0.2
0.1
C02
(%)
62.8
96.9
92.3
8.2
Novec™-612
(ppmv)
143.1
212.7
198.8
16.1
CO
(ppmv)
461.5
1,005.6
868.2
123.8
COF2
(ppmv)
28.5
35.9
32.3
2.3
HF
(ppmv)
132.7
450.3
381.2
79.2
CH,
(ppmv)
2.0
5.1
3.8
1.0
C2F6
(ppmv)
6.7
12.7
9.8
2.1
C2H4
(ppmv)
BDL
BDL
BDL
n/a
CH20
(ppmv)
BDL
0.4
0.2
0.1
CHF3
(ppmv)
4.1
7.4
6.6
0.8
N20
(ppmv)
2.8
4.7
3.5
0.5
N02
(ppmv)
0.4
1.7
0.9
0.4
C3F8
(ppmv)
6.3
21.2
13.5
5.2
Min (600ppmv)
Max (600ppmv)
Average (600ppmv)
Stdev
0.2
0.3
0.3
0.0
88.5
97.7
93.7
2.8
129.3
147.6
142.5
4.4
714.4
995.7
828.4
64.5
13.2
19.0
16.3
1.3
290.3
336.1
313.3
13.9
1.3
5.5
4.1
1.0
5.6
10.3
7.6
1.3
BDL
BDL
BDL
n/a
BDL
0.5
0.2
0.1
4.7
6.0
5.3
0.3
2.4
3.6
3.1
0.3
0.4
1.7
0.9
0.4
2.8
14.1
7.6
3.4
Min (400ppmv)
Max (400ppmv)
Average (400ppmv)
Stdev
0.2
1.5
0.4
0.4
34.5
96.9
86.3
16.1
31.9
77.3
55.4
9.2
100.8
1,286.1
838.9
287.7
BDL
6.5
2.9
1.7
82.7
180.0
110.9
21.7
1.5
19.4
5.1
4.8
BDL
4.8
1.9
1.1
BDL
3.8
0.6
1.2
BDL
3.1
0.5
0.8
0.4
4.3
2.4
0.8
2.2
6.3
3.2
1.0
0.7
5.3
1.6
1.3
BDL
3.0
0.7
1.6
Min (300ppmv)
Max (300ppmv)
Average (SOOppmv)
Stdev
0.2
1.3
0.5
0.4
49.1
95.3
81.6
17.5
2.2
3.7
3.2
0.5
451.3
5,031.0
1,871.0
1,361.1
BDL
BDL
BDL
n/a
6.9
23.6
11.9
5.3
3.1
10.6
5.6
2.5
BDL
BDL
BDL
n/a
BDL
BDL
BDL
n/a
BDL
1.0
0.4
0.2
0.1
0.4
0.3
0.1
1.1
3.0
2.2
0.6
1.0
3.8
1.9
1.0
BDL
BDL
BDL
n/a
Novec™612 Testing on 8/23/07
Pre Dross Min (200ppmv)
Pre Dross Max (200ppmv)
Pre Dross Avg. (200ppmv)
Stdev
0.2
0.9
0.4
0.2
63.4
98.0
89.0
10.2
65.5
96.9
75.8
7.8
438.9
1,287.1
764.7
211.2
3.5
8.9
6.3
1.3
146.0
244.8
170.0
23.6
BDL
67.3
8.9
17.4
1.6
5.8
3.6
1.5
BDL
19.9
2.3
5.3
BDL
10.3
1.4
2.9
2.5
4.8
3.6
0.7
2.1
4.8
2.9
0.7
BDL
2.1
0.5
0.5
BDL
7.9
3.3
3.5
Post Dross Min (200ppmv)
Post Dross Max (200ppmv)
Post Dross Avg. (200ppmv)
Stdev
0.2
0.9
0.3
0.1
68.4
96.9
92.1
5.3
59.8
81.8
64.2
3.8
852.5
1,983.2
1,421.0
227.4
2.0
6.1
4.9
0.8
24.9
160.0
120.8
28.0
1.4
7.6
4.8
1.3
0.7
1.8
1.2
0.3
BDL
BDL
BDL
n/a
BDL
0.4
0.2
0.1
1.9
3.3
2.6
0.3
2.0
4.1
2.8
0.5
1.4
5.0
2.3
0.8
BDL
BDL
BDL
n/a
Min (ISOppmv)
Max (ISOppmv)
Average (ISOppmv)
Stdev
0.2
1.1
0.3
0.1
61.7
98.7
91.7
6.4
24.7
40.3
30.8
2.5
677.2
2,028.2
1,270.8
267.2
BDL
1.1
0.4
0.5
46.3
117.0
61.3
13.0
2.4
12.6
4.9
1.9
0.1
0.9
0.6
0.1
BDL
2.5
0.4
0.5
BDL
2.4
0.3
0.4
0.8
2.3
1.5
0.2
1.6
3.4
2.3
0.4
BDL
3.0
1.0
0.6
BDL
BDL
BDL
n/a
BDL = below detectable limit; n/a = not applicable
4-3
-------
Table 4-2. MDL Summary for all Monitored Compounds During MTG-Shield™ Tests
Compound
H20
C02
Novec™-612
CO
COF2
HF
CH4
C2F6
C2H2
C2H4
CF4
CH202
CH2O
CH3F
CHF3
N2O
NO2
NO
OF2
SF6
C3F8
Minimum Detection
Limit (ppmv)
n/a
n/a
n/a
n/a
n/a
n/a
0.023*
0.104
4.08
0.379
3.803
0.256
0.312
1.333
0.022*
n/a
0.149
0.689
19.292
0.069
1.08
n/a= not applicable since compound was consistently present
throughout the testing
*Noise based MDL
4-4
-------
However, upon decreasing to 300 ppmv, a large increase in CO was observed. Referring
to the CO plot in Appendix A, a large CO spike followed by subsequent decay is observed which
is due to reaction of CO2 with the fresh Mg melt surface created by ingot loading. This may be
an indicating factor that the optimal cover gas concentration is near 300 ppmv for Novec™ 612.
A similar effect was observed after the dross performed during the 200 ppmv Novec™ 612
testing. Although a burn upon a clean melt surface should be expected, the CO concentrations
did not decrease throughout the post-dross 200 ppmv and 150 ppmv cover gas injections. Water
vapor was also observed throughout the entire test periods, mostly from ambient air dilution.
There were also absorbance features for two unknown compounds observed sporadically during
testing. One of these features was most likely due to a C-F or HC=CH functional group.
4.2 AM-Cover™ using HFC-134a with CDA Carrier Gas
Two concentrations of HFC-134a mixed with compressed dry air (CDA) were injected
(4,200 and 3,600 ppmv) during August 24th and 25th 2006. The HFC-134a cover gas standard
was diluted with facility CDA and mixed with a gas blender provided by Lunt Manufacturing
and operated by Advanced Magnesium Technologies (AMT). Each cover gas mixture's HFC-
134a concentration was verified by FTIR prior to injection into the crucible. To accomplish this,
the outlet of the gas blender was connected directly to the heated extraction line leading to the
FTIR. The gas blending parameters were adjusted until the desired HFC-134a concentration was
achieved. The gas mixture was then delivered into the crucible. Table 4-3 presents the data
summary for the observed compounds at each level tested during normal casting conditions.
Additional compounds were monitored but not observed above their respective MDLs. Table 4-
4 presents the MDLs for all the compounds monitored during the HFC-134a testing. In addition
to the compounds listed in Table 4-4, there was an unknown absorbance profile observed in the
HFC-134a spectra. Searches of URS reference libraries did not provide a positive match;
however, based on the absorbance frequencies and band shapes it is believed to be due to a
compound terminating with a double bonded hydrocarbon functional group (-CH=CH2). Time
series plots for the observed compounds are displayed in Appendix A.
4-5
-------
Table 4-3. Data Summary for AM-Cover™ using HFC-134a with CDA Carrier Gas
H2O
(%)
C02
(%)
CO
(ppmv)
HFC 134a
(ppmv)
HF
(ppmv)
COF2
(ppmv)
CH4
(ppmv)
C2H4
(ppmv)
CH20
(ppmv)
SF6
(ppmv)
N20
(ppmv)
N02
(ppmv)
NO
(ppmv)
SiF4
(ppmv)
AM-Cover Testing on 8/24/06
Min (4,200 ppmv)
Max (4,200 ppmv)
Average (4,200 ppmv)
Stdev
0.5
0.9
0.6
0.1
0.1
0.3
0.3
0.1
339.9
905.4
628.0
150.4
671.6
1,957.8
1,198.3
437.7
448.0
1,199.4
914.9
206.9
15.7
55.1
38.7
10.5
BDL
2.5
0.6
0.5
BDL
BDL
BDL
n/a
BDL
0.9
0.2
0.2
0.3
1.7
0.8
0.3
2.2
35.0
15.8
10.2
5.2
111
52.0
25.2
BDL
20.4
9.1
7.5
0.9
4.5
2.6
1.0
AM-Cover Testing on 8/25/06
Min (3,600 ppmv)
Max (3,600 ppmv)
Average (3,600 ppmv)
Stdev
0.5
1.1
0.6
0.1
0.2
0.3
0.3
0.0
311.9
713.0
452.2
106.3
475.7
1,512.3
810.7
276.1
669.8
929.5
785.2
58.9
16.1
59.1
37.3
11.8
BDL
2.7
0.5
0.5
BDL
BDL
BDL
n/a
BDL
0.6
0.1
0.1
0.2
0.8
0.3
0.1
4.7
38.7
19.9
7.1
22.7
84.5
66.6
12.1
0.8
28.0
18.5
8.3
0.3
1.4
0.6
0.3
BDL = below detectable limit
n/a = not applicable
4-6
-------
Table 4-4. MDL Summary for All Monitored Compounds During AM-Cover™ Tests
Compound
H20
C02
CO
HFC-134a
HF
COF2
CH4
C2H2
C2H4
CF4
CH202
CH2O
C2F6
CH3F
CHF3
SF6
OF2
N2O
NO2
NO
SiF4
Minimum Detection
Limit (ppmv)
n/a
n/a
n/a
n/a
n/a
n/a
0.374
0.163
0.213
3.80
11.99
0.162
0.566
0.916
0.278
n/a
0.676
n/a
n/a
1.99
n/a
n/a = not applicable since compound was consistently present
throughout the testing
The amount of HFC-134a destruction during each test was 71% and 77% for the 4,200
ppmv and 3,600 ppmv cover gas concentrations, respectively. It was evident from the HFC-134a
plots in Appendix A that the HFC-134a crucible concentrations decayed throughout the first two
hours of each test. Since there was no indication of magnesium burns during testing, the
injection of lower (optimized) HFC-134a cover gas concentrations may be acceptable and might
also provide slightly higher destruction rates and reduced emissions.
Gradual concentration decays were observed for CO and COF2. An inverse effect to
these decays was observed for HF and the nitrogen oxides (TSkO, NO, and NO2). This is easily
explained for HF, since the greater HFC-134a destruction in the cover gas produces a larger
concentration of the by-products. Over the first 2-hours of each test the NO2 concentration
continually increased before settling at approximately 70 ppmv. However, the N2O and NO
concentrations continually increased throughout each test. Although it was an interesting
dynamic, no plausible explanation could be inferred from the current data.
4-7
-------
In addition to unreacted HFC-134a and the nitrogen oxides formed from the CDA carrier
gas, destruction by-products (HF, COF2, and CO) were also observed during each test. HF
concentrations were measured from 448 to 1,199 ppmv with higher values correlating to higher
concentrations of F£FC-134a in the feed gas. Low concentrations of C2FL; and CH^O were also
sporadically observed, usually coincident with ingot loading. Detection of SF6 was also
observed at low concentrations and is a residual from previously used SF6 cover gas. SiF4 was
continuously observed and is believed to have formed as a result of contamination inside the
crucible, perhaps from fiberglass insulation. Water and CO2 were observed and were likely
present in the CDA carrier gas and any ambient air that entered the crucible through leaks.
4.3 SF6 with CDA Carrier Gas
Sulfur hexafluoride is the most commonly used cover gas in the magnesium die casting
industry, and exclusively used at Lunt Manufacturing. The SF6 cover gas was only monitored at
a single injection concentration (-3,000 ppmv) corresponding to the current levels used by Lunt
Manufacturing. The testing was completed on the 24th of August 2006 over two independent
periods. These periods were separated by five hours, the time when testing of the 4,200 ppmv
F£FC-134a cover gas occurred. A summary of all the observed compounds is presented in Table
4-5. Additional compounds were monitored, but not observed above their respective MDLs.
Table 4-6 presents the MDLs for all compounds monitored during the SF6 testing. Time series
plots for the observed compounds are displayed in Appendix A. The destruction of SFe during
testing was measured at approximately 33%. This is much lower than the other cover gases with
the exception of SO2 and frozen CO2 (note that SFe carries more available free fluorine per
molecule than does HFC-134a, and has less possible destruction-to-byproduct pathways than
doesNovec™612).
One observation of note was that the nitrogen oxides and FTP were observed at much
higher concentrations upon initial change over to SF6 from the HFC-134a. The limited data set
does show indication of decay and it is assumed that the concentrations would have continued to
the levels observed prior to the HFC-134a tests. The differences are believed to be an effect of
instability inside the crucible from temperature effects and/or the extended time it takes to
completely displace the previously used cover gas. Another factor in this effect may involve the
reaction dynamics between the molten magnesium and the SFe/CDA mixture.
4-8
-------
Table 4-5. Data Summary for SF6 with CDA Carrier Gas
H2O
(%)
CO2
(%)
CO
(ppmv)
SF6
(ppmv)
HF
(ppmv)
CH,
(ppmv)
C2H2
(ppmv)
CH2O
(ppmv)
N2O
(ppmv)
NO2
(ppmv)
NO
(ppmv)
SO2
(ppmv)
SF6 Testing on 8/24/06 (first run)
Min (3,000 ppmv)
Max (3,000 ppmv)
Average (3,000 ppmv)
Stdev
0.9
2.1
1.2
0.2
0.0
0.1
0.0
0.0
5.0
183.2
23.0
33.7
641.8
2,187.1
1,966.1
325.9
0.7
8.2
3.5
1.8
BDL
32.2
3.1
4.9
BDL
2.8
0.4
0.8
BDL
7.0
0.7
0.9
0.9
4.1
2.4
0.5
3.5
8.1
6.0
1.0
BDL
BDL
BDL
n/a
2.3
56.4
35.9
11.9
SF6 Testing on 8/24/06 (second run)
Min (3,000 ppmv)
Max (3,000 ppmv)
Average (3,000 ppmv)
Stdev
1.0
1.2
1.1
0.1
0.0
0.1
0.0
0.0
3.5
46.8
11.1
13.7
1,760.6
2,032.4
1,932.2
97.8
27.1
49.1
34.0
6.7
0.9
5.5
3.0
1.5
BDL
BDL
BDL
n/a
BDL
1.0
0.3
0.3
17.6
32.2
23.1
3.7
33.5
47.1
38.2
4.2
2.1
15.9
7.4
3.7
23.2
51.9
36.6
9.1
BDL = below detectable limit
n/a = not applicable
4-9
-------
Table 4-6. MDL Summary for All Monitored Compounds During SF6 Tests
Compound
H2O
C02
CO
SF6
HF
COF2
CH4
C2H2
C2H4
CF4
CH2O2
CH2O
C2F6
CH3F
CHF3
NO
OF2
N2O
NO2
NO
SO2
Minimum Detection Limit (ppmv)
n/a
n/a
n/a
n/a
n/a
1.02
0.230
0.039*
41.00
0.013
0.280
0.204*
0.290
0.303
0.129
1.27
0.934
n/a
0.107
1.27
1.71
n/a = not Applicable since compound was consistently present
throughout the testing.
*Noise based MDL
Although there is no conclusive evidence, the theory is that with a clean melt surface,
created from a dross or during an ingot drop, the probability to form HF is maximized from
greater amounts of liberated fluorine (as the SF6 is destroyed via reaction with the melt surface).
As the protective MgO/MgF layer forms atop the melt surface, it acts as a barrier separating the
melt surface from the SF6/CDA cover gas, thereby minimizing the destruction of SF6 and
subsequent formation of FTF.
SF6 destruction to form FtF and SO2 was also observed throughout the tests. The SO2
concentrations were rather variable ranging from 2-56 ppmv during the tests. Sharp decreases
were observed during ingot loading, presumably from dilution. FTF concentrations, on average,
were 4 and 34 ppmv for the first and second tests, respectively. The recovery of the SO2 back to
the pre-ingot loading concentration trended behind the recovery time of the other gases. As
previously mentioned nitrogen oxides (N2O and NO2) were observed throughout the test and are
likely resultant from the destruction of the CD A carrier gas. Formaldehyde and acetylene were
also detected during the tests. This was the only cover gas where acetylene (C2H2) was
observed. Ambient air components (CH4, CO2, H2O and CO) were observed throughout the
tests. With the exception of CO, all the averaged ambient air compounds were near or slightly
4-10
-------
above normal concentrations. CO had transient spikes on the order of 47-183 ppmv presumably
from short-lived burns inside the crucible.
4.4 SO2 with CDA Carrier Gas
Sulfur dioxide mixtures were injected at six different concentrations over a two-day
period beginning on the 28th of August and concluding on the 29th of August, 2006. The gas
mixer was built and operated by Polycontrols. The existing CDA supply was passed through a
desiccant-type dryer to remove excess moisture prior to blending with SC>2. An initial SC>2
concentration of 1% was injected followed by successively lower concentration injections to a
minimum concentration of 0.4% SC>2. Two sampling tests at 0.5% SC>2 were conducted
approximately 16 hours apart. A summary of all the observed compounds is presented in Table
4-7. The MDLs for all the compounds monitored during the SC>2 tests are listed in Table 4-8 and
time series plots of concentration data are displayed in Appendix A.
There were few destruction byproducts that could be attributed to 862. Most of the
observed compounds were either a carry-over effect of residual chemicals from the previous
cover gas used in the crucible (e.g., SF6 and HF), ambient air components (H2O, CC>2, CH4),
destruction byproducts formed from ambient air dilution during ingot loading (CH^O and C2H/t)
or nitrogen oxides formed from the CDA carrier gas. Slightly elevated levels of observed CO,
with respect to ambient levels, were also observed. H2SO4 was not measured at concentrations
above its minimum detection limit (0.051 ppmv) within the crucible headspace.
Five different SO2 concentrations (1%, 0.85%, 0.7%, 0.6% and 0.5%) were injected
during the first day of SO2 testing followed by two concentrations on the second day (0.5% and
0.4%). As the 862 concentration was lowered, the relative rate of destruction also decreased. A
repeat of the 0.5% SO2 test was done at the beginning of the 2nd day of SO2 testing and the
destruction rates were similar, 30% during day 1 and 37% during day 2. However, the overall
SO2 destruction throughout the entire testing only varied marginally (29% - 37%). This small
variance is indicative of an equilibrium effect between the SO2/CDA cover gas mixture and the
liquid melt as compared to the other mixtures, which appeared to rely more on total mass
coverage of the melt surface. In other words, the carrier gas (CDA) most likely plays a more
prominent role in melt coverage than do the diluents for the fluorinated cover gases, which rely
more on the amount of available F mass to produce MgF coverage. The least amount of 862
destruction (29%) was observed during the 0.6% SC>2 concentration and the largest amount
destroyed was 37% during the 1% and 0.5% 862 cover gas concentrations.
4-11
-------
Table 4-7. Data Summary for SO2 with CDA Carrier Gas
H20
(%)
C02
(ppmv)
CO
(ppmv)
CH,
(ppmv)
S02
(ppmv)
N20
(ppmv)
N02
(ppmv)
NO
(ppmv)
CH20
(ppmv)
C2H,
(ppmv)
SF6
(ppmv)
HF
(ppmv)
SO2 Testing on 8/28/06
Min (1%)
Max (1%)
Average (1%)
Stdev (1%)
0.3
0.6
0.4
0.1
409.5
492.5
435.2
28.6
1.8
17.1
6.0
6.3
1.0
6.5
2.7
1.9
5,280.7
6,617.5
6,086.6
390.2
2.9
3.9
3.4
0.4
1.1
2.2
1.5
0.4
0.8
2.5
1.5
0.7
BDL
0.3
0.2
0.7
BDL
1.7
0.4
1.9
0.1
0.2
0.1
0.0
3.6
8.9
6.8
2.1
Min (0.85%)
Max (0.85%)
Average (0.85%)
Stdev (0.85%)
0.2
0.7
0.3
0.1
398.4
484.7
421.6
25.4
1.0
7.2
2.0
2.1
1.2
3.2
1.9
0.6
4,277.8
5,933.5
5,521.3
463.9
2.6
3.4
2.8
0.2
0.9
1.8
1.3
0.3
0.5
1.5
1.0
0.3
BDL
BDL
BDL
n/a
BDL
0.3
0.1
0.1
0.0
0.1
0.1
0.0
6.5
12.9
8.7
2.0
Min (0.70%)
Max (0.70%)
Average (0.70%)
Stdev (0.70%)
0.2
0.6
0.3
0.1
401.0
547.4
424.9
40.2
0.9
16.4
3.0
4.0
1.3
5.4
2.3
1.0
3,555.4
5,041.9
4,652.5
370.6
2.3
3.6
2.9
0.3
0.8
2.1
1.3
0.3
0.6
1.5
1.1
0.3
BDL
BDL
BDL
n/a
BDL
0.4
0.1
0.1
0.0
0.1
0.0
0.0
3.9
11.5
6.1
2.2
Min (0.60%)
Max (0.60%)
Average (0.60%)
Stdev (0.60%)
0.2
0.4
0.2
0.1
394.7
448.2
406.5
12.5
0.8
17.3
2.4
3.6
1.5
4.0
2.2
0.5
3,563.8
4,380.7
4,157.3
256.2
2.3
3.3
3.0
0.3
0.4
1.3
1.1
0.2
0.5
1.4
1.0
0.2
BDL
1.1
0.2
0.3
BDL
1.4
0.1
0.3
0.0
0.1
0.0
0.0
3.2
7.4
4.1
1.0
Min (0.50%)
Max (0.50%)
Average (0.50%)
Stdev (0.50%)
0.1
0.4
0.2
0.1
392.1
438.9
406.3
15.4
0.9
7.3
2.3
1.9
1.5
3.3
2.1
0.6
2,788.3
3,689.3
3,401.3
311.2
2.4
3.4
2.9
0.2
0.4
1.1
0.9
0.2
0.4
1.1
0.8
0.2
BDL
0.4
0.2
0.2
BDL
0.7
0.1
0.2
0.0
0.1
0.0
0.0
2.3
6.1
3.6
1.2
SO2 Testing on 8/29/06
Min (0.50%)
Max (0.50%)
Average (0.50%)
Stdev (0.50%)
0.3
0.8
0.4
0.1
428.1
494.4
442.5
15.5
1.4
33.1
5.0
7.3
1.0
8.8
2.3
1.8
1,980.9
3,284.3
3,042.0
300.2
1.6
2.4
1.9
0.2
0.8
1.5
1.2
0.2
BDL
BDL
BDL
0.0
BDL
0.9
0.3
0.2
BDL
3.5
0.3
0.7
0.0
0.1
0.1
0.0
0.1
0.3
0.2
0.0
Min (0.40%)
Max (0.40%)
Average (0.40%)
Stdev (0.40%)
0.3
0.7
0.4
0.1
410.0
479.9
432.1
14.3
1.1
22.3
3.7
4.1
1.0
5.2
2.0
1.1
1,890.8
2,717.7
2,518.2
196.5
1.4
2.0
1.6
0.1
0.5
1.3
1.0
0.2
BDL
BDL
BDL
n/a
BDL
0.5
0.3
0.2
BDL
1.1
0.2
0.3
0.0
0.1
0.0
0.0
0.1
0.3
0.2
0.1
BDL = below detectable limit; n/a = not applicable
4-12
-------
Table 4-8. MDL Summary for All Monitored Compounds During SO2 Tests
Compound
H20
CO2
CO
CH4
S02
HF
SF6
N20
NO2
NO
CH20
C2H4
H2SO4
S03
Minimum Detection Limit
(ppmv)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
0.335
0.213
0.075
0.051
9.81
n/a = not applicable since compound was consistently present
throughout the testing.
Another interesting trend was observed in the HF data between both 0.5% SC>2 tests.
Measured HF concentrations ranged from 0.1 ppmv to 13 ppmv over both tests. The HF
concentrations during the second test were much lower (95% less) than what was observed during
the 0.5% SC>2 test on the first day, presumably due to continuing displacement of free fluorine off
the melt surface and complete purging of residual SFe cover gas out of the crucible environment.
4.5 Frozen CO2
The final cover gas tested was frozen CC>2. Testing began at 1651 on the 29th of August and
ended at 1828 of the same day. The frozen CC>2 was formed by expansion of compressed
liquid/gaseous CC>2. As the gas cooled through expansion, it froze, forming solid phase CC>2 which
was gravity fed into the crucible. The obvious difference between the frozen CC>2 and all the other
cover gases is the method of injection and its chemical state. Compressed CC>2 was delivered to the
nozzle at a pressure of 100 pounds per square inch (psi) from a liquid CC>2 dewar. The
delivery/formation of the frozen CC>2 was controlled and optimized by regulating the injection into
the crucible with a solenoid valve.
4-13
-------
The amount of CC>2 destruction was minimal averaging 4% over the entire sample period.
This was the lowest destruction of all the cover gases. As a result, CC>2 had very few
decomposition by-products, with the main one being CO. A small amount of C2H4 was consistently
present throughout the testing. Sulfur dioxide, HF and SFe were observed during the testing, but
were likely present as a carry-over effect of residual chemicals from previous cover gases used in
the crucible. The presence of nitrogen oxides was most likely due to thermal decomposition of
ambient air entering the crucible through leaks and intrusion during ingot loading. Table 4-9 lists
the observed compounds during the frozen CC>2 cover gas testing. Table 4-10 presents all the
compounds monitored along with the MDLs for those that were not observed. Time series plots for
each observed compound during the frozen CC>2 cover gas test are presented in Appendix A.
4-14
-------
Table 4-9. Data Summary for Frozen CO2
Min (100%)
Max (100%)
Average (100%)
Stdev
H2O
(%)
0.1
0.6
0.2
0.1
CO2
(%)
52.7
100.0
95.7
11.5
CO
(ppmv)
181.9
883.9
401.8
202.0
SO2
(ppmv)
9.2
37.4
22.5
5.9
HF
(ppmv)
BDL
0.2
0.0
0.0
CH4
(ppmv)
BDL
2.6
0.8
0.7
C2H,
(ppmv)
0.6
1.3
0.8
0.2
N2O
(ppmv)
BDL
3.4
0.5
0.9
NO2
(ppmv)
BDL
0.9
0.1
0.2
NO
(ppmv)
BDL
1.7
0.6
0.6
SF6
(ppmv)
0.1
0.4
0.2
0.1
BDL = below detectable limit
n/a = not applicable
4-15
-------
Table 4-10. MDL Summary for All Monitored Compounds During Frozen CO2 Testing
Compound
H2O
C02
CO
SO2
HF
CH4
C2H2
C2H4
CH2O2
CH2O
COF2
N20
NO2
NO
SF6
Minimum Detection Limit
(ppmv)
n/a
n/a
n/a
n/a
n/a
n/a
0.287
n/a
0.253
0.193
0.069
0.466
0.117
1.03
n/a
n/a = not applicable since compound was consistently present
throughout the testing.
4.6 Determination of Dilution
As discussed at the end of Section 3.1 and throughout 3.2, quadrapole mass spectrometry
(QMS) was used to monitor N2 inside the crucible for a determination of dilution (via ambient air).
With the exception of MTG-Shield™ using Novec™ 612 and frozen CC>2, all the cover gases used
CDA as their carrier gas. Since ambient air is the only source of dilution into the crucible, it was
impossible to determine dilution from all the cover gases where CDA was used. Furthermore, since
frozen CO2 was gravity fed to the crucible as a solid it was unique and not representative for all the
other cover gases. Therefore, dilution was determined from the MTG-Shield™ cover gas tests and
applied to all the other cover gas tests, except frozen CC>2. Since the sampling system was a closed
system and the Novec™ 612 cover gas was diluted only with CC>2, any N2 present was from the
intrusion of ambient air into the crucible headspace. A slip-stream of the extracted sample gas was
delivered to the QMS and monitored for N2. The QMS response was calibrated for nitrogen over a
concentration 0%-50% range using the Polycontrols gas blender. This calibration was then fit to a
linear regression (y = ax + b) and applied to the sample data using Equation 4-1. Figure 4-1
displays the QMS calibration curve.
y=10.65e-9x + 2.86e-9
Equation 4-1
Where:
y = N2 response as read by the QMS and
x = % N2
4-16
-------
o
«
&
6.00E-07 -,
5.00E-07 -
4.00E-07 -
3.00E-07 -
2.00E-07 -
1.00E-07 -
O.OOE + 00
y = 1.065E-08X + 2.863E-09
R2 = 9.996E-01
0.00 10.00 20.00 30.00 40.00
Percent Nitrogen
50.00
60.00
Figure 4-1. QMS Response Curve for 0-50% Nitrogen
The a and b coefficients in the calibration equation (Equation 4-1) are on the order of 10"9
because they represent the QMS detector response, which was an electron multiplier that produced
electrical signals in the nanoamp (10"9 amperes) range. Nitrogen was monitored for the entire
MTG-Shield testing performed on the 23rd of August and the raw data is plotted in Figure 4-2. The
average QMS N2 response was then applied to Equation 4-1 giving the % N2 observed in the
crucible. Since ambient air is 78% N2 the dilution was calculated from Equation 4-2.
Yo Dilution = % N2 / 0.78 (as calculated from Equation 4-1)
Equation 4-2
Applying equations 4-1 and 4-2 to the QMS data set gave an average dilution of 3.0%
during the MTG-Shield™ testing. For the purposes of this study dilution is considered to be the
amount of ambient air intrusion resulting in a proportional decrease in head-space cover gas
concentration relative to the feed gas concentration. This dilution set included data when the ingot
door was open for loading and is therefore only applicable to this particular die-casting machine
fabricating this particular part. The dilution was also calculated with the ingot loading periods
omitted. Surprisingly, at 2.7% dilution, there is little difference between the two, thus indicating
that the majority of the 3.0% dilution is from leaks in the crucible lid allowing for ambient air
ingression into the headspace. Oxygen was also monitored continuously over a two-hour period
during the MTG-Shield™ testing as described in Section 3.1. The periods where the ingot door was
opened were omitted and the average 62 concentration inside the crucible was calculated at 4.5%.
This was slightly greater than the 2.7% observed using the QMS monitoring of N2 when the ingot
4-17
-------
loading periods were omitted. The QMS and C>2 analyzers were not running simultaneously which
may be a cause for the slight discrepancy. It should be noted that it is virtually impossible to
measure the actual crucible dilution with complete accuracy due to limitations of the operational
setting. Dilution during the frozen CC>2 testing was estimated to be 0.4% using the N2 tracer
method - this is likely due to the high volume of CC>2 being injected into the crucible.
Estimated dilution rates of 3% - 4.5% are significantly lower than what was estimated
during previous research in 2003. Dilution in the 2003 study was estimated to be on the order of
20% - 25% using intermittent measurements of C>2 as the only dilution indicator.8 The significant
reduction in dilution is likely due to the following factors:
• The crucible in this study was significantly "tighter" due to the addition of insulation
batting that sealed all openings for potential ambient air intrusion;
• The rate of ingot loading (about every 15 minutes) was l/5th that of the previous study
resulting in significantly less ambient air intrusion during this process; and
• The use of continuous measurements of N2 as a dilution indicator provided a more
accurate (and lower) estimate than the previous approach.
6.00E-08
-------
4.7 Occupational Exposure Monitoring
Each of the cover gases evaluated in this study can produce emissions that may be of
concern from an occupational exposure standpoint. For this reason, a second FTIR was used to
monitor ambient air for any potential occupational exposure hazards associated with the usage of
each cover gas. For example, formaldehyde (CH2O), carbonyl fluoride (COF2), SO2 and FTP have
very low Occupational Safety and Health Administration (OSHA) 8-hour time-weighted average
permissible exposure limits (PELs) of 0.75, 2, 2, and 3 ppmv, respectively.9 Occupational exposure
details, including short term exposure limits (STELs) and symptoms of chronic exposure for the
primary compounds and byproducts of concern during monitoring are provided in Table 4-11. As
an additional point of interest, in prior measurement trials involving Novec™ 612, the presence of
perfluoroisobutylene (PFIB) as a possible byproduct of cover gas degradation was noted; however,
monitoring for this compound was not possible due to the absence of an available spectral FTIR
reference.10
Two worker breathing zones were monitored during the testing. The first zone was
monitored during the use of the MTG-Shield™, AM-Cover™, and SF6 cover gases. This zone was
located at the end of the process where the part was robotically dropped to a worker who inspects
and files it. This is the area where the worker operating the machine spent the bulk of his time.
The 2nd zone was at the crucible lid near the ingot door and was monitored during the SO2 and
frozen CO2 cover gases. This zone represents a worst-case location in that it is near the most
probable location for direct exposure to cover gas emissions. This is especially relevant to the use
of SO2 and the likelihood for there to be direct exposure to cover gas during crucible maintenance
operations such as dressing.
It should be noted that this study should be considered only as a screening of potential risks
from an occupational perspective. A true industrial hygiene occupational exposure study would
involve personal air-packs worn by workers during normal operations. However, this study does
provide highly accurate information resulting from continuous monitoring of ambient air at the
primary worker stations.
9 OSHA Permissible Exposure Limits (PELs).
10 Mibrath, D., "Development of 3M™ Novec™ 612 Magnesium Protection Fluid as a Substitute for SF6 Over Molten
Magnesium," International Conference on SF6 and the Environment: Emission Reduction Technologies, November 21-
22,2002, San Diego, CA.
4-19
-------
Table 4-11. Cover Gas Compound Occupational Exposure Details
Compound
SFfi- Sulfur
Hexafluoride—
HFC-134a- 1,1,1.2
Tetrafluoroethane—
Novec™ 61212
CO - Carbon
Monoxide—
HF - Hydrofluoric AcidM
SO, - Sulfur Dioxide15
COF2 - Carbonvl
Fluoride12
CH2O - Formaldehyde12
PEL
1,000
1,000
150
50
3.0
2.0
2.0
0.75
STEL
NA
NA
NA
400
6.0
5.0
5.0
2.0
Acute Exposure
Nausea, vomiting, difficulty breathing, dizziness, fatigue,
emotional disturbances, tingling sensation, suffocation,
convulsions, coma
Rapid evaporation of the liquid may cause frostbite. The
substance may cause effects on the central nervous system and
cardiovascular system, resulting in cardiac disorders.
Contact with the eyes during product use is not expected to result
in significant irritation. Contact with the skin or ingestion during
product use is not expected to result in significant irritation or
health effects. May present an inhalation hazard if thermal
decomposition occurs. The primary thermal decomposition
byproducts of concern include CO, HF, CO2, and
perfluoroisobutylene (PFIB).
Acute exposure to carbon monoxide may include headache,
flushing, nausea, vertigo, weakness, irritability, unconsciousness,
and in persons with pre-existing heart disease and
atherosclerosis, chest pain and leg pain.
The substance is corrosive to the eyes, the skin and the
respiratory tract. Inhalation of this gas or vapor may cause lung
oedema. The substance may cause hypocalcemia. Exposure
above the OEL may result in death.
Respiratory tract irritation, rhinorrhea, choking, and coughing.
Irritation to eyes, skin, mucous membrane, respiratory system;
eye, skin burns; lacrimation (discharge of tears); cough,
pulmonary edema, dyspnea (breathing difficulty), frostbite.
Irritation eyes, nose, throat, respiratory system; lacrimation
(discharge of tears); cough; wheezing.
Chronic Exposure
No information available.
No information available.
No information available.
Repeated bouts of carbon
monoxide poisoning may
cause anorexia, headache,
lassitude, dizziness, and
ataxia.
The substance may cause
fluorosis.
Permanent pulmonary
impairment, which is caused
by repeated episodes of
bronchoconstriction.
Gastrointestinal pain,
muscle fibrosis, skeletal
fluorosis.
Potential occupational
carcinogen.
11 Matheson Tri Gas. Material Safety Data Sheet: Sulfur Hexafluoride. 2003.
12ICSC: NENG 1281 International Chemical Safety Cards (WHO/IPCS/ILO), CDC/NIOSH: HFC 134a, 1998.
13 3M Material Safety Data Sheet, 3M™ Novec™ 612 Magnesium Protection Fluid, 2007.
14 Tennessee OSHA. Instruction CPL 2.501 - Local Emphasis Program - Carbon Monoxide, 1999.
15 ICSC: NENG 0283 International Chemical Safety Cards (WHO/IPCS/ILO), CDC/NIOSH: HF, 2000.
16 CDC/NIOSH 1988 OSHA PEL Project Documentation: Sulfur Dioxide, 1998.
17 NIOSH Publication No. 2005-151. NIOSH Pocket Guide to Chemical Hazards: Carbonyl Fluoride. 2005.
18 NIOSH Publication No. 2005-151. NIOSH Pocket Guide to Chemical Hazards: Formaldehyde. 2005.
4-20
-------
4.7.1 Observed Compounds During Occupational Exposure Monitoring
With the exception of one instance during the use of 862, monitoring at both sampling
zones did not detect concentrations of any cover gas compound or byproduct that would be of
concern from an occupational exposure perspective. The primary compounds of concern were
either monitored at levels that were well below their respective PELs, or were below detectable
limits of the FTIR. Table 4-12 lists the all the primary compounds of concern monitored during the
testing.
SFe was continuously observed throughout the testing regardless of the cover gas being used
for melt protection. A continuous background level of SF6 was expected since it is used as a cover
gas in all crucibles throughout the facility. Carbon monoxide was also continuously observed well
above ambient levels regardless of location or cover gas. This was most likely due to combustion
during the short-lived burns that occurred as each shot of molten magnesium is cast and/or possible
CO formation as a byproduct during small burns within the crucibles. All the observed compounds
were below the OSHA 8-hour PEL. The one incidence that indicated the potential for an
occupational exposure concern occurred during the use of 862. A maximum recorded reading of
1.6 ppmv was observed which is approaching the OSHA 8-hour PEL of 2 ppmv. This brief
occurrence was observed approximately 30 inches from, and 12 inches above the ingot door during
a period when it was open for a prolonged period of time due to an ingot loading malfunction. This
measurement identifies a health concern to the worker that would occur while performing
maintenance on the crucible lid or performing a dross. It is certain that personal protective
equipment (PPE) in the form of a respirator would be required for workers performing crucible
maintenance on a machine running dilute SO2 as a cover gas.
While these tests do not represent a rigorous industrial hygiene analysis, the results illustrate
that given the ventilation present at this facility, the crucible head space gases are sufficiently
diluted to an extent that harmful concentrations were not found in ambient air close to the primary
worker breathing zones. The potential for occupational exposure at the ground level of the facility
is also reduced due to the convection currents carrying escaping cover gas upward towards the
ceiling and the high-volume exhaust fans.
4-21
-------
Table 4-12. Monitored Compounds from Occupational Exposure Ambient Monitoring
Cover Gas
Date
Zone**
Permissible Exposure Limit (PEL)
Short-term Exposure Limit (STEL)
MTG-Shield™
8/22
1
Max
Average
Novec™612
(ppmv)
150
n/a
BDL
BDL
HFC-134a
(ppmv)
1000
n/a
n/a
n/a
SF6
(ppmv)
1000
n/a
0.08
0.06
S02
(ppmv)
2
5
n/a
n/a
HF
(ppmv)
3
6
BDL
BDL
CO
(ppmv)
50
400
11.60
7.07
COF2
(ppmv)
2
5
BDL
BDL
CH20
(ppmv)
0.75
2
BDL
BDL
MTG-Shield™
8/23
1
Max
Average
BDL
BDL
n/a
n/a
0.19
0.06
n/a
n/a
BDL
BDL
14.45
6.80
BDL
BDL
BDL
BDL
AM-Cover™
8/24
1
Max
Average
n/a
n/a
0.050
0.020
0.16
0.10
n/a
n/a
BDL
BDL
15.33
5.28
BDL
BDL
BDL
BDL
AM-Cover™
8/25
1
Max
Average
n/a
n/a
0.078
0.038
0.29
0.18
n/a
n/a
BDL
BDL
6.48
2.75
BDL
BDL
BDL
BDL
SF6
8/24
1
Max
Average
n/a
n/a
n/a
n/a
0.28
0.18
BDL
BDL
BDL
BDL
9.48
2.75
BDL
BDL
BDL
BDL
S02
8/28
2
Max
Average
n/a
n/a
n/a
n/a
0.23
0.19
1.6*
0.14
BDL
BDL
7.22
1.21
BDL
BDL
BDL
BDL
SO2
8/29
2
Max
Average
n/a
n/a
n/a
n/a
0.07
0.03
BDL
BDL
BDL
BDL
11.49
7.06
BDL
BDL
BDL
BDL
Frozen CO2
8/29
2
Max
Average
n/a
n/a
n/a
n/a
0.23
0.16
n/a
n/a
BDL
BDL
11.46
4.55
BDL
BDL
BDL
BDL
BDL = below detectable limit
n/a = not applicable
*Occurred during an instance when the ingot loading door was open for a prolonged period of time due to a loading malfunction.
"Zone 1 was located at the process end where the part is robotically dropped to the worker. Zone 2 was located near the ingot loading area of the
crucible.
4-22
-------
5.0 Conclusions
5.1 Cover Gas Test Observations
The cover gas test results were described in detail in Section 4 and will be summarized in
this section. Each cover gas was sampled, in an identical fashion, from the crucible head space
of Lunt Manufacturing's HPM 400 die-casting machine (#1) fabricating an automotive lock body
housing. Furthermore, all the cover gases except frozen CC>2 were delivered to the crucible head
space in a similar manner. This testing setup allowed for a direct comparison of five different
magnesium melt cover gases as they relate to greenhouse gas emissions and occupational
exposure.
One observation which occurred multiple times throughout the testing was the
inconsistencies between measurements using the same cover gas performed on different days.
This was most prominent in the HF, Novec™ 612 and nitrogen oxide concentrations during the
MTG-Shield™ and SF6 tests. The most likely cause was instability in the crucible brought on by
a cover gas change. It is evident from constant gradual decays observed in the cover gas
concentrations upon initial injection, as well as the length of time required to effectively purge
the carry-over gases from previous cover gases, that there was a significant time period
associated with purging the crucible and reaching equilibrium in the head-space.
5.1.1 MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas
The primary destruction by-products detected when using MTG-Shield™ using Novec™
612 gas were; CO, COF2, HF, C^f^ CHFs and CsFg. Measurements were conducted while
successively lowering concentrations of Novec™ 612 cover gas beginning at 800 ppmv and
ending at 150 ppmv. The first day of testing started with the Novec™ 612 cover gas injected at a
concentration of 800 ppmv and lowered to 600, 400 and 300 ppmv. As the Novec™ 612
concentration decreased, COF2, HF, C2F6, CHF3 and C3F8 concentrations also decreased in
succession although not by the same percentages as the cover gas. The low Novec™ 612, HF
and other by-products observed at the end of the first day of sampling (COF2 and C$6 were
BDL and Novec™ 612 and HF averaged less than 5 and 12 ppmv respectively) did not continue
the downward trend during the 2nd day even though lower cover gas concentrations (200 and 150
ppmv) were used. The 300 ppmv Novec™ 612 cover gas concentration used during the 1st day
of testing resulted in lower HF, Novec™ 612 and other by-products than either the 200 or 150
ppmv Novec™ 612 cover gas concentrations tested during the 2nd day. It is believed that the
length of Novec™ 612 usage during the first day of testing provided a more stable crucible
operation toward the end of the 6-hour testing episode than the testing conducted on day 2.
5-1
-------
5.1.2 AM-Cover™ using HFC-134a with CDA Carrier Gas
The primary destruction by-products detected when using AM-Cover™ using HFC-134a
cover gas were CO, COF2, and HF. Measurements were conducted using two HFC-134a cover
gas concentrations (4,200 and 3,600 ppmv). This 14.8% decrease in cover gas concentration had
very little effect on COF2 reduction (4%) but did reduce the HFC-134a, HF, and CO
concentrations in the crucible by 32%, 14% and 28%, respectively. Concentrations of these
compounds steadily decayed throughout the tests indicating that upon a cover gas change there
was a significant period required to effectively purge out the crucible headspace. Additional
byproducts such as CH2O, NO, N2O and NO2 and C2H4 were formed, or slightly increased, with
the addition of ambient air during the ingot loads. The non-process gases, SF6 and SiF4 were
also observed throughout testing. The SFe presence was a residual from its use prior to AM-
Cover™.
5.1.3 SF6 with CDA Carrier Gas
A 3,000 ppmv SFe/CDA cover gas mixture was monitored over two independent time
periods. The primary compounds detected when using SF6 are: HF and SO2. For both cold
chamber measurements, average HF concentrations were on the order of 4 to 34 ppmv. Low
levels of NO, N2O and NO2 were detected, ranging from 1 to 47 ppmv. CH2O and C2H2 were
observed sporadically throughout testing near their respective FTIR detection limits. This was
the only cover gas where C2H2 was observed. Ambient air components (CH4, CO2, H2O and
CO) were observed throughout the testing. With the exception of CO, all the averaged ambient
air compounds were near or slightly above their atmospheric concentrations. CO had transient
spikes on the order of 47 - 183 ppmv, presumably from short-lived burns inside the crucible.
5.1.4 SO2 with CDA Carrier Gas
Sulfur dioxide cover gas was monitored over a 2-day period at six different
concentrations. The SO2 cover gas concentrations ranged from a high of 1% to a low of 0.4 %.
There were no destruction by-products which could be solely attributed to SO2. The remainder
of the observed compounds are either a function of residual gases from the previous cover gas
used in the crucible (SF6 and HF), ambient air components (H2O, CO2, CH4), by-products
formed from ambient air dilution during ingot loading (CH2O and C2H4) or nitrogen oxides
formed from the CDA carrier gas and/or ambient air dilution from ingot loading. Ambient
concentrations of SO2 were below occupational exposure limits or not detectible during the study
(See Section 4.7 for more details).19
19 All applicable safety precautions (e.g., operational procedures) should be followed when using SO2.
5-2
-------
5.1.5 Frozen CO2
Frozen CC>2 was the final cover gas tested. The frozen CO2 was gravity fed to the
crucible as a solid and sublimed once inside. As with SO2, CO2 had very few decomposition by-
products with the main one being CO. A very small amount of C2H4 was consistently present
throughout the testing. SC>2, HF and SFe were observed during the testing but were carry-over
gases from previous cover gases used in the crucible. The presence of nitrogen oxides is most
likely due to thermal decomposition of ambient air entering the crucible through leaks and ingot
loading. The amount of CC>2 destroyed during the testing averaged 3%. This was by far the
lowest amount destroyed of all the cover gases tested. Since the injection flow rates were not
accurately monitored and the CC>2 dewar was not weighed before and after the tests, the absolute
amount of CC>2 used was unknown.
5.2 Cover Gas Destruction
One of the main objectives with this cover gas study was to determine the level of
destruction. Destruction estimates were calculated as:
_. _„ . Delivery Concentration - Dilution Corrected Measured Concentration
Percent Destruction =
Delivery Concentration
Table 5-1 provides a summary of all the tests and calculated destruction rates. The
measured concentration value in the above equation was corrected for dilution due to leaks and
ingot loading as calculated by the QMS.20
20 See section 4.6 for further explanation of dilution determination
5-3
-------
Table 5-1. Percent Destruction for Cover Gas Testing
Cover Gas Mixture
Components
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/C02
Novec™612/C02
Novec™612/C02
HFC-134a/CDA
HFC-134a/CDA
SF6/CDA
SF6/CDA
S02/CDA
S02/CDA
S02/CDA
S02/CDA
SO2/CDA
SO2/CDA
SO2/CDA
Frozen CO2
Table
4-1
4-1
4-1
4-1
4-1
4-1
4-1
4-3
4-3
4-5
4-5
4-7
4-7
4-7
4-7
4-7
4-7
4-7
4-9
Date
8/22/06
8/22/06
8/22/06
8/22/06
8/23/06
8/23/06
8/23/06
8/24/06
8/25/06
8/24/06
8/24/06
8/28/06
8/28/06
8/28/06
8/28/06
8/28/06
8/29/06
8/29/06
8/29/06
Time
0910-1005
1008-1135
1138-1301
1347-1425
0906-0958
1108-1248
1252-1556
1510-1736
0810-1215
0939-1311
1806-1902
1105-1130
1133-1218
1222-1348
1351-1500
1504-1552
0845-1001
1005-1155
1651-1829
Cover Gas
Mixture Flow a
(1pm)
-36
-36
-36
-36
-36
-36
-36
-40
-40
-35
-35
-39
-39
-39
-39
-39
-39
-39
Dewar at 100 psi
Cover Gas
Delivery Cone. a
(ppmv)
800
600
400
300
200*
200**
150
4,200
3,600
3,000
3,000
10,000
8,500
7,000
6,000
5,000
5,000
4,000
1,000,000
Cover Gas
Measured Cone.
(ppmv)
198.8
142.5
55.4
3.2
75.8
64.2
30.8
1,198.3
810.7
1,966.1
1,932.2
6,086.6
5,521.3
4,652.5
4,157.3
3,401.3
3,042.0
2,518.2
957,390.1
Dilution Factor
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.00
Cover Gas
Destruction
74%
76%
86%
99%
61%
67%
79%
71%
77%
32%
34%
37%
33%
32%
29%
30%
37%
35%
4%
aAs provided by Lunt Manufacturing,
*Data collected before a dross.
**Data collected after a dross.
AMI, Poly controls, and Matheson Tri-Gas/TNSC.
5-4
-------
The percent destruction of Novec™ 612 increased as the cover gas concentration was
successively lowered and exhibited the highest destruction rate of all the cover gases tested,
ranging from 61% - 99%. A concentration between 150 and 400 ppmv appears to be optimal
once the crucible has been sufficiently purged with the Novec™ 612 cover gas. Destruction of
HFC-134a also increased as the concentration of the cover gas was lowered. HFC-134a was not
optimized to the extent that Novec™ 612 was, so only limited conclusions can be drawn when
comparing the two. It is believed that the destruction would have increased had lower HFC-134a
cover gas concentrations been tested. As it was, the HFC-134a displayed the 2nd highest
destruction (71% - 77%). SF6 was tested on two occasions at the same concentration and was
consistent with respect to amount destroyed. However, at approximately 33% it was much less
than the cover gas destruction rates for Novec™ 612 and HFC-134a (>61% or greater). The SC>2
cover gas destruction was not significantly dependant on injected concentration (to the extent
tested). Like SF6: the percent destroyed was relatively low, at 29% - 37%. Finally, CC>2
displayed the lowest destruction at only 4%.
The destruction rates estimated for SF6 in this study were significantly higher than what
was estimated during the 2003 study. This is likely due to the much lower feed gas
concentrations utilized in this study - 3,000 ppmv compared to 5,000 and 19,000. Also, the
reduced level of dilution in this study results in the destruction of the cover gas having a larger
share of the reduction in measured concentration.
5.3 Global Climate Change Potential Discussion
One of the benefits of using AM-Cover™, MTG-Shield™, 862 or CC>2 as cover gases for
magnesium melt protection is their contribution to global climate change is significantly lower
when compared to SF6. This is evident when comparing their estimated global warming
potentials (GWPs). Table 5-2 presents GWPs of several compounds detected during this study.
Table 5-2. Comparison of 100-Year GWP Estimates from the
Intergovernmental Panel on Climate Change (IPCC)
Second (1996) Assessment Report
Gas
Methane
Nitrous Oxide
HFC-134a
Perfluoromethane (CF4)
Perfluoroethane (C2F6)
Perfluoropropane (C3F8)
Sulfur Hexafluoride (SF6)
IPCC GWP
21
310
1,300
6,500
9,200
7,000
23,900
IPCC (1996), Climate Change 1995: The Scientific of Climate Change.
Intergovernmental Panel on Climate Change, Cambridge University Press.
Cambridge, U.K.
5-5
-------
The crucible headspace contains a large variety of compounds, but only those with
corresponding GWP values were used in estimating the overall GWP impact of switching to
alternate cover gases. This calculation consisted of multiplying the average concentrations (parts
per million by volume) for each of the component cover gases and applicable destruction
products, with their respective GWP factors (obtained from the Second Assessment Report of the
Intergovernmental Panel on Climate Change) to obtain a GWP -weighted value. The summation
of all the GWP -weighted values for a particular cover gas resulted in the normalized CC>2
equivalent which was compared to the CC>2 equivalent corresponding to
Table 5-3 shows that when comparing the normalized CC>2 equivalent, or composite
GWP, the alternate cover gases have a much lower impact. An obvious source for this reduction
can be found in a direct comparison of cover gas GWPs shown in Table 5-2. Novec™ 612's
GWP has not been supplied by the IPCC, but is likely to be extremely low (i.e., Novec™ 612 is
a fluorinated ketone, which is assumed to have a GWP on the order of I).22 In addition to having
lower GWPs, the alternate fluorinated cover gas compounds have much higher destructions (on
the order of 61-99%) as compared to SF6 (on the order of 33%). While the use of HFC-134a and
Novec™ 612 does produce destruction byproducts with GWPs, their impact is minimal due to
the very low concentrations generated. Although the SC>2 and CC>2 cover gases degrade as much
or less than SF6, they both have much lower GWPs than SF6. Compared to using SF6, switching
to AM-Cover™ or MTG-Shield™ produces a reduction in overall global warming impact of at
least 97%. 23 Changing the cover gas from SF6 to dilute 862 reduces the global warming impact
even more (>99.9%) but introduces a more complex operational scenario due to toxicity
concerns. Switching to frozen CC>2 provides a reduction in global warming impact of 98%
relative to SFe.
The above comparison does not include the specific flow rates for each cover gas. In
order to provide a more comprehensive analysis of composite GWP, an additional comparison
was conducted. Using the ideal gas law, the molecular weights of each gas and the delivery flow
rate of the cover gas was used to estimate the composite emission rate in grams per hour (g/hr).
This equation can be described as follows:
21 Residual SF6 concentrations were excluded from the overall GWP calculations for the alternative gases.
22 D'Anna B, Sellevag S.R., Wirtz K., and Nielsen C.J. Photolysis Study of Perfluoro-2-methyl-3-pentanone Under
Natural Sunlight Conditions Environ Sci and Tech 2005 39(22) 8708-8711
23 Please refer to Section 5.4 for a discussion regarding the uncertainty associated with this methodology.
5-6
-------
Emission Rate = ppm x MW x Ipm x ^-(38.6 liters I mole x 106)
^ hour ) hour
ppm = measured average concentration in parts per million
MW = molecular weight in grams per mole
1pm = gas flow in liters per minute
These values were summed to provide a composite GWP value that was weighted by the
cover gas flow rate. The average flow weighted GWP values were then compared against the
corresponding values for SF6. Based on this approach, all of the cover gas alternatives were
observed to reduce GHG emissions by at least 98% relative to SFe.24 Details of the flow-
weighted GHG emission impacts are presented in Table 5-4.
5.4 Uncertainty Discussion
The results of this measurement study should not be interpreted to represent an absolute
analysis of GHG emissions associated with HFC-134a, Novec™ 612, 862, CC>2 and SF6 cover
gas usage. While this study does present a relatively accurate measurement analysis and
approximate comparison of GHG emissions, there are several areas of uncertainty inherent with
this methodology. These areas of uncertainty include FTIR and QMS error, error associated with
blending gases, dilution correction, and analytical and operational variation of the die-casting
machine evaluated.
Measurements taken by the FTIR and QMS are subject to variability inherent with highly
complex analytical equipment. While all prudent steps were taken during the measurement study
to minimize this contributor to uncertainty (see Sections 3.3 and 4.6), a small degree of error is
unavoidable.
24 It should be noted that fully optimized cover gas systems will result in slightly better performance than the
average values reported here.
5-7
-------
Table 5-3. Normalized GWP Comparison of Measured Emissions from Inside the Crucible Headspace
Table
4-1
4-1
4-1
4-1
4-1
4-1
4-1
4-3
4-3
4-5
4-5
4-7
4-7
4-7
4-7
4-7
4-7
4-7
4-9
Cover Gas
Mixture
Components3
Novec™612/CO2
Novec™612/C02
Novec™612/C02
Novec™612/C02
Novec™612/C02
Novec™612/CO2
Novec™612/CO2
HFC-134a/CDA
HFC-134a/CDA
SF6/CDA
SF6/CDA
S02/CDA
SO2/CDA
SO2/CDA
SO2/CDA
SO2/CDA
S02/CDA
S02/CDA
Frozen CO2
Cover Gas
Delivery
Cone, (ppm)
800
600
400
300
200
200
150
4,200
3,600
3,000
3,000
10,000
8,500
7,000
6,000
5,000
5,000
4,000
1,000,000
GWP
Weighted
Cover Gasb
199
143
55
3
76
64
31
1,557,833
1,053,924
46,989,729
46,180,244
0
0
0
0
0
0
0
0
GWP
Weighted
C02
951,127
964,973
889,151
841,046
916,790
949,285
944,754
2,781
2,520
468
486
435
422
425
407
406
443
432
957,390
GWP
Weighted
CH,
80
85
106
118
188
101
102
12
11
66
65
57
40
47
46
45
47
43
17
GWP
Weighted
N20
1,098
951
991
688
896
868
714
4,896
6,180
751
7,173
1,039
871
886
929
894
597
508
169
GWP
Weighted
C2F6
89,864
69,697
17,221
0
33,380
10,947
5,191
0
0
0
0
0
0
0
0
0
0
0
0
GWP
Weighted
C3F8
94,430
53,260
4,940
0
23,368
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Normalized
CO2
Equivalent
1,136,798
1,089,109
912,465
841,855
974,697
961,265
950,792
1,583,642
1,070,841
46,991,015
46,187,968
1,532
1,333
1,359
1,381
1,345
1,087
983
957,576
Average by
Cover Gas
980,997 c
1, 327,242 d
46,589,491 e
l,288f
957,576 g
Chg.
from
SF6
98%
97%
99.9%
98%
a As provided by Lunt Manufacturing, AMT, Polycontrols, and Matheson Tri-Gas/TNSC.
b GWP weighting based on dilution corrected concentration for the primary cover gas constituent (e.g., NovecT
"Average composite GWP forNovec™ 612/CO2tests (Table 4-1)
d Average composite GWP for HFC-134a/CDA tests (Table 4-3)
e SF6 composite GWP baseline estimate for comparison with other tests (Table 4-5)
f Average composite GWP for SO2 test (Table 4-7)
g Average composite GWP for CO2 test (Table 4-9)
'612,HFC-134a, SF6)
5-S
-------
Table 5-4. GWP (Weighted by Cover Gas Flow) Comparison of Measured Emissions from Inside the Crucible Head Space
Table
4-1
4-1
4-1
4-1
4-1
4-1
4-1
4-3
4-3
4-5
4-5
4-7
4-7
4-7
4-7
4-7
4-7
4-7
4-9
Cover Gas
Mixture
Components3
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/CO2
Novec™612/C02
Novec™612/C02
Novec™612/C02
HFC-134a/CDA
HFC-134a/CDA
SF6/CDA
SF6/CDA
S02/CDA
S02/CDA
S02/CDA
S02/CDA
SO2/CDA
SO2/CDA
SO2/CDA
Frozen CO2
Cover Gas
Delivery
Cone, (ppm)
800
600
400
300
200
200
150
4,200
3,600
3,000
3,000
10,000
8,500
7,000
6,000
5,000
5,000
4,000
1,000,000
GWP
Weighted
Cover Gasb
(g/hr)
4
3
1
0
1
1
1
9,882
6,686
373,336
366,905
0
0
0
0
0
0
0
0
GWP
Weighted
C02
(g/hr)
2,342
2,377
2,190
2,071
2,258
2,338
2,327
8
7
1
1
1
1
1
1
1
1
1
8,460
GWP
Weighted
CH,
(g/hr)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
GWP
Weighted
N20
(g/hr)
3
2
2
2
2
2
2
13
17
2
17
3
2
2
2
2
2
1
0
GWP
Weighted
C2F6
(g/hr)
694
538
133
0
258
85
40
0
0
0
0
0
0
0
0
0
0
0
0
GWP
Weighted
C3F8
(g/hr)
994
560
52
0
246
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Normalized
CO2 GWP
Equivalent
(g/hr)
4,037
3,480
2,378
2,073
2,765
2,426
2,369
9,903
6,710
373,339
366,923
4
3
4
4
4
3
3
8,460
Average
by Cover
Gas
(g/hr)
2,790C
8,306 d
370,131 e
3f
8,460g
Chg.
from
SF6
99%
98%
>99.9%
98%
a As provided by Lunt Manufacturing, AMI, Polycontrols, and Matheson Tri-Gas/TNSC.
bGWP weighting based on dilution corrected concentration for the primary cover gas constituent (e.g., Novec™ 612, HFC-134a, SF6)
"Average composite flow rate weighted GWP for Novec™ 612/CO2tests (Table 4-1)
d Average composite flow rate weighted GWP for HFC-134a/CDA tests (Table 4-3)
e SF6 composite flow rate weighted GWP baseline estimate for comparison with other tests (Table 4-5)
f Average composite flow rate weighted GWP for SO2 test (Table 4-7)
g Flow rate weighted GWP for CO2 is based on an estimated flow rate of 0.0012 gal/second when solenoid is open.
5-9
-------
Characterization of Emissions and Occupational Exposure
Associated with Melt Protection Technologies for Magnesium
Die Casting
Appendix A
October 2007
EPA 430-R-07-008
-------
Table of Contents
Page
A-l MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas 3
A-2 AM-Cover™ using HFC-134a with CDA Carrier Gas 18
A-3 SF6 with CDA Carrier Gas 35
A-4 Dilute SO2 with CDA Carrier Gas 42
A-5 Frozen CO2 48
-------
Appendix A-1
MTG-Shield™ using Novec™ 612 with CO2 Carrier Gas
Plots
-------
TM
Time Series Plot for Novec-612 on 8/22/06
250 n
05." 05: 05.
300 ppm Injection
•Novec-612
/>*> K> K^ ^
Time (hh:mm:ss)
-------
Time Series Plot for Novec™-612 on 8/23/06
120 n
100 --
E
Q.
80 -
C
o
15 so H
•+J
0)
o
C
O 40 -[
20 --
Lowered Novec Concentratation
From 200ppm to 150 ppm
•Novec 8/23
Time (hh:mm:ss)
-------
TM
Time Series Plot for PFCs Using Novec -612 Cover Gas on 8/22/06
XXXXXXXXXXXXX * .... .Y_Y \.r.r.r.-r.
05." <$.- 05." 05.
Time (hh:mm:ss)
-------
TM
Time Series Plot for PFCs Using Novec -612 Cover Gas on 8/23/06
E
Q.
a.
c
o
+J
2
*->
0)
u
c
o
o
10
9
Lowered Novec Concentratation
From 200ppm to 150 ppm
XXX X XXXXXXXXXXXXXXXXXXXXXXXXXXXX
Time (hh:mm:ss)
-------
TM
Time Series Plot for HF Using Novec -612 Cover Gas on 8/22/06
300 ppm Injection
A1"
Time (hh:mm:ss)
-------
TM
Time Series Plot for HF Using Novec -612 Cover Gas on 8/23/06
300 -
250 --
200 --
Q.
a.
c
o
IS 150 H
•4-1
0)
o
c
O 100
50 -
Dross
Lowered Novec Concentratation
From 200ppm to 150 ppm
0 --
r^ r^ oo oo
CM ro ^ in
CM CM
00 05
i- CM
O>
CO
o o
in o
•<- CM CM CM
CM ro ^ in
CD CD CD CD CD CD
Time (hh:mm:ss)
-------
TM
Time Series Plot for CO2 and CO Using Novec -612 Cover Gas on 8/22/06
6000
120
Time (hh:mm:ss)
10
-------
TM
Time Series Plot for CO2 and CO Using Novec -612 Cover Gas on 8/23/06
2500
2000
500
Dross
Lowered Novec Concentratation
From 200ppm to 150 ppm
-- 100
-- 80
120
-- 60
-- 20
VK> K^'K^'T^'T^^
Time (hh:mm:ss)
c
o
0)
u
c
o
o
40 O
u
-------
TM
Time Series Plot for Nitrogen Oxides Using Novec -612 Cover Gas on 8/22/06
E
Q.
a.
c
o
0)
u
c
o
o
05." 05." 05." 05.
^^'^'^'K^'K> K^'K^'K^'K^' ^' **' $'' ^' ^'' •^'^'^'^'^
Time (hh:mm:ss)
12
-------
TM
Time Series Plot for Nitrogen Oxides Using Novec -612 Cover Gas on 8/23/06
5 -
?4
Lowered Novec Concentratation
. From 200ppm to 150 ppm
CD
o
i^ i^ co co
CM ro •t in
T-cNCNrororo'^-'^-inincDCD
CNro^moT-CNro'^-inoT-
O) O) O) O) O) O)
Time (hh:mm:ss)
13
-------
TM
Time Series Plot for CH2O and C2H4 Using Novec -612 Cover Gas on 8/22/06
4.5
3.5 -
Q.
a.
IT 2.5
o
+J
2
•£ 2
0)
o
c
O 1.5
0.5
800 ppm Injection
AA
600 ppm Injection
400 ppm Injection 300 ppm Injection
A
AA
Time (hh:mm:ss)
14
-------
TM
Time Series Plot for CH2O and C2H4 on Using Novec -612 Cover Gas 8/23/06
E
Q.
C
o
0)
u
C
o
o
25
20
15
10
I
Dross
Lowered Novec Concentratation
From 200ppm to 150 ppm
Time (hh:mm:ss)
15
-------
TM
Time Series Plot for CH4 and H2O Using Novec -612 Cover Gas on 8/22/06
25
E
Q.
a.
c
o
+j
ra
8
o
o
Tt
I
O
10
5 -
.<$\&
?- - - - " - - - -
Time (hh:mm:ss)
16
-------
TM
Time Series Plot for CH4 and H2O Using Novec -612 Cover Gas on 8/23/06
E
Q.
a.
c
o
0)
u
c
o
o
Tt
I
O
Lowered Novec Concentratation
From 200ppm to 150 ppm
-- 1
1.2
,, ,, ,, ,,
Time (hh:mm:ss)
17
-------
Appendix A-2
AM-Cover™ using HFC-134a with CDA Carrier Gas
Plots
18
-------
Time Series Plot for HFC-134a During Injection at 4200 ppm on 8/24/06
2500
•HFC134a
A*
A
Time (hh:mm:ss)
19
-------
Time Series Plot for HFC-134a During Injection at 3600 ppm on 8/25/06
1600
1400
-HFC1343
Or
.
Q,-
Time (hh:mm:ss)
20
-------
Time Series Plot for COF2 During HFC-134a Injection at 4200 ppm on 8/24/06
E
Q.
a.
c
o
U-*
2
•4-1
0)
o
c
o
o
20
10
A*
Time (hh:mm:ss)
21
-------
Time Series Plot for COF2 During HFC-134a Injection at 3600 ppm on 8/25/06
E
Q.
C
o
0)
o
C
o
o
10
Time (hh:mm:ss)
22
-------
E
Q.
a.
c
o
+J
2
*->
0)
u
c
o
o
Time Series Plot for Non-Process By-Products
During HFC-134a Injection at 4200 ppm on 8/24/06
fV ,-V .A" .A1" A
NT NT NT NT
Time (hh:mm:ss)
23
-------
0)
u
c
o
o
1.6
1.4
1.2
0. 1
a.
c
o
13 0.8
Time Series Plot for Non-Process By-Products
During HFC-134a Injection at 3600 ppm on 8/25/06
.<§>
-
)v^ )v^ )v^ )v^ N N
Time (hh:mm:ss)
.$>
>
. . . .
N N. K- N\- N\- N\-
24
-------
Time Series Plot for HF During HFC-134a Injection at 4200 ppm on 8/24/06
1400
^ AQ ,.& J&
N*y
-------
Time Series Plot for HF During HFC-134a Injection at 3600 ppm on 8/25/06
1000
900
c
o
ts 50°
+j
o 400
c
o
o
300
200
100
NK: N*'
Time (hh:mm:ss)
26
-------
Time Series Plot for CO During HFC-134a Injection at 4200 ppm on 8/24/06
1000
900
fo nN A ,> .0 -fo rA <* A <•> A
•V .-V .-^ .A- A~ x.Cf ^.O
Time (hh:mm:ss)
27
-------
Time Series Plot for CO During HFC-134a Injection at 3600 ppm on 8/25/06
800
Time (hh:mm:ss)
28
-------
E
Q.
C
o
*->
2
•+j
0)
u
C
o
o
0 4i
Time Series Plot for Nitrogen Oxide By-Products
During HFC-134a Injection at 4200 ppm on 8/24/06
A* AQ ,.<$>
^•^' «& ^ „«£• «#' «&' «#*£*& <#' +& <&' <#'
\\ \\ \\ \\
A•
V
Time (hh:mm:ss)
29
-------
Time Series Plot for Nitrogen Oxide By-Products
During HFC-134a Injection at 3600 ppm on 8/25/06
E
Q.
a.
c
o
+J
ra
*->
0)
u
c
o
o
30
20
10
0 4
<&• <&• <&• <&• <&• °r
-------
Time Series Plot for CH2O and C2H4 During HFC-134a Injection at 4200 ppm on 8/24/06
^ V^ VV V^ Vv • ^' ^' *& *& *&
Time (hh:mm:ss)
A'V A'- A'-" A'-" A'-
V V V V V
31
-------
Time Series Plot for CH2O During HFC-134a Injection at 3600 ppm on 8/25/06
Time (hh:mm:ss)
32
-------
Time Series Plot for Ambient Air Compounds
During HFC-134a Injection at 4200 ppm on 8/24/06
NVD- NVD- NVD- NVD- NVD
Time (hh:mm:ss)
33
-------
Time Series Plot for Ambient Air Compounds
During HFC-134a Injection at 3600 ppm on 8/25/06
. . .
Q,- Q,- Q, Q, Q,-
Time (hh:mm:ss)
34
-------
Appendix A-3
SF6with CDA Carrier Gas
Plots
35
-------
Time Series Plot of SF6 During SF6 Injection at 3000ppm on 8/24/06
2500
Time (hh:mm:ss)
36
-------
Time Series Plot of HF During SF6 Injection at 3000ppm on 8/24/06
E
Q.
C
o
0)
o
C
o
o
20
10
Time (hh:mm:ss)
37
-------
Time Series Plot of SO2 During SF6 Injection at 3000ppm on 8/24/06
Time (hh:mm:ss)
38
-------
Time Series Plot of N2O and NO2 During SF6 Injection at 3000ppm on 8/24/06
E
Q.
C
o
IE
*->
0)
u
C
o
o
J$'.*
Time (hh:mm:ss)
39
-------
Time Series Plot of C2H2 and CH2O During SF6 Injection at 3000ppm on 8/24/06
I
a.
c
o
4
0)
o
o
.
o
r.& f& efS>
-
s§>
O' O NT NT NT NT NT ND ND N°
Time (hh:mm:ss)
40
-------
Time Series Plot of Ambient Compounds During SF6 Injection at SOOOppm on 8/24/06
200
c
o
2 1.5
0)
u
c
o
o
CM
O
O
O
CM
I
I- 0
^ -
Time (hh:mm:ss)
41
-------
Appendix A-4
Dilute SO2 with CDA Carrier Gas
Plots
42
-------
Time Series Plot for SO2 During SO2 Cover Gas Injections on 8/28-8/29, 2006
7000
Time (hh:mm:ss)
43
-------
Time Series Plot for SF6 and HF During SO2 Cover Gas Injections on 8/28-8/29, 2006
0.2
0.18
E
Q.
C
o
0)
u
C
o
o
Time (hh:mm:ss)
44
-------
Time Series Plot for Nitrogen Oxides During SO2 Cover Gas Injections on 8/28-8/29, 2006
MHHHHIHHHHHII MHHHHIHHHHHHIHHHHI
A ^ ,> n
-------
Time Series Plot for C2H4 and CH2O During SO2 Cover Gas Injections on 8/28-8/29, 2006
3.5
E
Q.
C
o
IE
*->
0)
u
C
o
o
v s$> n? .*r .<§> &..$'
Time (hh:mm:ss)
46
-------
Time Series Plot for Ambient Compounds During
SO2 Cover Gas Injections on 8/28-8/29, 2006
o
in T-
^ ro
in oo r^
o CM ro
Rl 5
" s
OO CD
in
^ o T- CM T- CM
3 o
CD m
ro in
in oo o r^
CM ^ o T-
5
in ro o r^
o CM ^ m
_. oo
^ ^
8
ro
CM
o r^
^ in
CM CM CM 00 00 00
^^^^101010000)0)0)0)
r^ CD m •t
•=r o r\i •=!;
•t^ O) CD OO
T- r\i •=!; o
000-^
OO CM CM
o r\i '^r
T- oo in
CM oo in
Time (hh:mm:ss)
47
-------
Appendix A-5
Frozen CO2 Plots
48
-------
c
o
0)
o
c
o
o
120
100
A*-
V V
N«-
^ J*
?• &
&
Time (hh:mm:ss)
49
-------
Time Series Plot of CO During CO2 Injection on 8/29/06
E
Q.
a.
c
o
U-*
2
•+j
0)
o
c
o
o
1000
900
800
700
600
500
Time (hh:mm:ss)
50
-------
Time Series Plot of SO2 and SF6 During CO2 Injection on 8/29/06
E
Q.
a.
c
o
0)
u
c
o
o
CM
O
40
35
^ ^ jP jP ^ $* ^ f? jp ^ jy ^ ^
Time (hh:mm:ss)
51
-------
Time Series Plot of Nitrogen Oxides During CO2 Injection on 8/29/06
3.5
E
Q.
c
0
0)
U
O
O
2.e
ie
n , rl
Time (hh:mm:ss)
52
-------
Time Series Plot of C2H4 During CO2 Injection on 8/29/06
E
Q.
a.
c
o
0)
o
c
o
o
^ /'
xs
Time (hh:mm:ss)
53
-------
Time Series Plot of CH4 and H2O During CO2 Injection on 8/29/06
E
Q.
C
o
0)
u
C
o
o
Tt
o
A •
\V
A •
\V
<#' ^
\v \v
------- |