In-Depth Survey Report:
Control Technology Assessment
for the Welding Operations
at
Boilermaker's National Apprenticeship Training School
Kansas City, Kansas
REPORT WRITTEN BY:
Marjorie E. Wallace
Thomas Fischbach
Ronald J. Kovein
REPORT DATE:
June 27, 1997
REPORT NO.:
ECTB 214-13a
U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
Public Health Service
Centers for Disease Control and Prevention
National Institute for Occupational Safety and Health
Division of Physical Sciences and Engineering
4676 Columbia Parkway - R5
Cincinnati, Ohio 45226
PLANT SURVEYED:
Boilermaker's National Apprenticeship
Training School
1017 N. 9th Street
Kansas City, Kansas 66101
SIC CODE:
8249
SURVEY DATE:
April 22-26, 1996
SURVEY CONDUCTED BY:
Marjorie Edmonds Wallace
Ronald J. Kovein
Daniel S. Watkins
EMPLOYER REPRESENTATIVES
Louie Lombardi
CONTACTED:
Lead Welding Instructor
John Standish
Welding Instructor
Mike Blood
Welding Instructor
Paul Ross
Union Welder
Glenn Tubbs
Business Manager, Local 83
Roger Erickson
Business Representative, Lodge 83
OTHER REPRESENTATIVES
Pam Susi
CONTACTED:
Center to Protect Workers' Rights
Bob Dinsmore
Plymovent, Inc.
ANALYTICAL SERVICES:
DataChem Laboratories
Salt Lake City, Utah
MANUSCRIPT PREPARATION:
Bernice L. Clark
DISCLAIMER
Mention of company names or products does not constitute endorsement by the Centers for Disease Control and
Prevention.
The Engineering Control Technology Branch of the National Institute for Occupational Safety and Health is currently
conducting a study of welding operations and workers' exposures to welding fumes. The goal of this study is to
identify, observe, and evaluate engineering control measures which may reduce the amount of fume a worker is exposed
to during welding. At the conclusion of this study, information on effective control technology will be disseminated
to the welding community. This report summarizes the results of an in-depth sampling study conducted on the welding
operations at the Boilermaker's National Apprenticeship Training School. The goals of the study were to measure fume
exposures during stainless steel welding, using different rod types and rod diameters, and to evaluate the effect of
local exhaust ventilation during outdoor welding operations.
During this study, shielded metal arc (or stick) welding techniques were evaluated, primarily in a semi-enclosed tank
located outside the school. Four standard types of stainless steel rods used by the Boilermaker's were identified and
evaluated: AWS 308, 309-16, 316, and 347. The 3xx series designated by the American Welding Society (AWS)
have a high chromium-nickel makeup. Four standard rod diameters used by the Boilermaker's were also identified and
evaluated: 3/32", 1/8", 5/32", and 3/16". The rods were from a number of manufacturers, including
Alloy Rods, Harris Welco Alloys, McKay, Tech Alloy, and Lincoln. Two local exhaust ventilation units were evaluated
during this study, a mobile fume extractor with a 2 meter arm and a portable fan unit with a similar exhaust arm.
Both units were provided by Plymovent.
Gravimetric samples were collected and analyzed for total welding fume levels and for welding fume constituent
levels. Additional samples were collected for hexavalent chromium. Real-time data was collected on
relative concentrations using aerosol photometers and particle counters. Gas levels were determined for NO2,
CO and O2.
Results indicated that welders were overexposed to stainless steel welding fume, hexavalent chromium, arsenic, total
chromium, iron, manganese, and nickel. Local exhaust ventilation helped to reduce these levels but not to a point
where the exposures were considered to be completely controlled. The effect of wind and the position of the welder
may have been extremely detrimental to the ability of the local exhaust ventilation at controlling welding fume
exposures.
Over the past twenty years, the National Institute for Occupational Safety and Health (NIOSH) has recognized the
importance of preventing potential health hazards associated with fumes and gases generated during welding operations
(see Appendix A); however, no comprehensive study of control technology for welding operations has been conducted
since the late 70s. As such, the Engineering Control Technology Branch (ECTB) of NIOSH is currently conducting a
study to evaluate the effectiveness of engineering control measures in reducing welding fume exposures. This welding
assessment study was initiated for several reasons. First, even with advances in control technology, welders continue
to be exposed to hazardous welding fumes and gases.1 Second, the continual development and
implementation of new welding processes, techniques, and materials can result in unidentified and uncontrolled health
hazards. Third, many welding operations are small shops that may not have access to current technology for the
control of welding emissions; this project responds to the NIOSH small business initiative which identifies welding
shops as one of the top ten hazardous small businesses, in terms of occupational health risks.2
Finally, as it is likely that welding will be a high priority for OSHA over the next few years,3
industry will need timely research on engineering technology for the control of welding fumes and gases.
Many sites use a combination of ventilation and respiratory protection equipment to try and control the amount of
fumes (and gases) the welder is exposed to during welding operations. If the ventilation system does not adequately
control the fumes, the welder often relies heavily on the respirator for protection against potential health hazards.
Ideally, respiratory protection should be used only as a last resort against welding fumes and only when an
appropriate respiratory protection program is in place. It is unclear whether strong respiratory protection programs
are common in welding shops. Therefore, the goals of this assessment study are to identify effective ventilation
systems, or other engineering control measures, that will protect the welder's health, and to disseminate this
information to the welding community.
In January 1995, the Hazard Evaluation and Technical Assistance Branch (HETAB) of NIOSH received numerous requests
from building and construction trade unions requesting evaluations of stainless steel welding fumes. Aware of ECTB's
ongoing welding project, HETAB requested engineering assistance on the control of stainless steel welding fumes
during boiler rehab work. To determine exposure levels and effective controls, ECTB needed to evaluate various
systems and processes in the field. The Boilermakers' Union offered to participate in a simulation study where they
would provide welders and welding equipment and consumables at their training facility in Kansas City, Kansas. Two
types of portable local exhaust ventilation units, supplied by Plymovent (Mississauga, Ontario), were evaluated for
their ability to exhaust stainless steel welding fumes and gases away from the worker's breathing zone, at the point
of generation. The ventilation units were tested both inside a building and outside in a semi-enclosed
tank. Process variables, such as the welding rod type and diameter, were also evaluated to determine their effect on
welding fume levels.
The effect of welding fumes and gases on a welder's health can vary depending on such factors as the length and
intensity of the exposure and the specific toxic metals involved. Welding processes involving stainless steel,
cadmium- or lead-coated steel, or metals such as nickel, chrome, zinc, and copper are particularly hazardous as the
fumes produced are considerably more toxic than those encountered when welding mild steel. Mild steel consists mainly
of iron, carbon, and small amounts of manganese, phosphorous, sulfur, and silicon, while stainless steel contains
mainly iron, chromium, nickel, titanium, and manganese.4 The NIOSH criteria document
identifies arsenic, beryllium, cadmium, chromium (VI), and nickel as potential human carcinogens that may be present
in welding fumes. Epidemiological studies and case reports of workers exposed to welding emissions have shown an
excessive incidence of acute and chronic respiratory diseases. Welder respiratory ailments can include occupational
asthma, siderosis, emphysema, chronic bronchitis, fibrosis of the lung, and lung cancer. Epidemiological evidence
indicates that welders generally have a 40 percent increase in relative risk of developing lung cancer as a result of
their work.4 Other cancers associated with welding include leukemia, cancer of the stomach,
brain, nasal sinus, and pancreas. Cadmium poisoning can affect the respiratory system and damage the liver and
kidneys. A common reaction to overexposure to metal fumes, particularly zinc oxide fumes, is metal fume fever, with
symptoms resembling the flu. Other health hazards during welding can include vision problems and dermatitis arising
from ultraviolet radiation exposures, burns, and musculoskeletal stress from awkward work positions.4
See Appendix B for additional information on potential health hazards from welding.
As a guide when evaluating hazards posed by workplace exposures such as those from welding, NIOSH field staff employ
environmental evaluation criteria. These criteria are intended to suggest levels of exposure to which most workers
may be exposed up to 10 hours per day, 40 hours per week for a working lifetime without experiencing adverse health
effects. It is, however, important to note that not all workers will be protected from adverse health effects even if
their exposures are maintained below these levels. A small percentage may experience adverse health effects due to
individual susceptibility, a preexisting medical condition, and/or a hypersensitivity (allergy). In addition, some
hazardous substances may act in combination with other workplace exposures, the general environment, or with
medications or personal habits of the worker to produce health effects even if the occupational exposures are
controlled at the level set by evaluation criteria. These combined effects are often not considered in the evaluation
criteria. Also, some substances are absorbed by direct contact with the skin and mucous membranes, and thus
potentially increase the overall exposure. Finally, evaluation criteria may change over the years as new information
on the toxic effects of an agent become available.
The primary sources of environmental evaluation criteria in the United States that can be used for the workplace are:
(1) the U.S. Department of Labor (OSHA) Permissible Exposure Limits (PELs); (2) NIOSH Recommended Exposure Limits
(RELs); and (3) the American Conference of Governmental Industrial Hygienists's (ACGIH) Threshold Limit Values (TLVs).
The OSHA PELs are required to consider the feasibility of controlling exposures in various industries where the
agents are used; the NIOSH RELs, by contrast, are based primarily on concerns relating to the prevention of
occupational disease. ACGIH Threshold Limit Values (TLVs) refer to airborne concentrations of substances and
represent conditions under which it is believed that nearly all workers may be repeatedly exposed day after day
without adverse health effects. ACGIH states that the TLVs are guidelines. The ACGIH is a private, professional
society. It should be noted that industry is legally required to meet only those levels specified by OSHA PELs.
In 1989, the OSHA PEL for total welding fume was set at 5 mg/m3 (5000 µg/m3)
as an 8-hour time-weighted average (TWA); however, this limit was vacated and currently is not enforceable. Since
1989, OSHA has not reestablished a PEL for total welding fume; however, individual PELs have been set for the various
constituents which can be found in welding fumes (see Appendix C).5 OSHA has also set a PEL
for total particulate not otherwise regulated (PNOR) at 15 mg/m3 as an 8-hourtime-weighted
average (TWA). A TWA exposure refers to the average airborne concentration of a substance during a normal 8- to 10-hour
workday. Some substances have recommended short-term exposure limits (STEL) or ceiling values that are intended to
supplement the TWA where there are recognized toxic effects from high, short-term exposures.
The ACGIH has set a TLV-TWA for welding fumes-total particulate (NOC) at 5 mg/m3. The ACGIH
recommends that conclusions based on total fume concentration are generally adequate if no toxic elements are present
in the welding rod, metal, or metal coating and if conditions are not conducive to the formation of toxic gases.6
NIOSH indicates that it is not possible to establish an exposure limit for total welding emissions since the
composition of welding fumes and gases vary greatly, and the welding constituents may interact to produce adverse
health effects. Therefore, NIOSH suggests that the exposure limits set for each welding fume constituent should be
met (see Appendix C). However, it was noted in the NIOSH criteria document4 that even when
welding fume constituents were below the PELs, there was still excesses in morbidity and mortality among welders. As
such, NIOSH recommends that welding emissions should be controlled with current exposure limits considered to be
upper limits.4
This study was conducted at the National Apprenticeship Training School for the International Brotherhood of
Boilermakers. At this facility, union apprentices undergo 144 hours of classwork, over a four year period, as part of
their training program. Welding is a primary focus of the training; boilermaker apprentices must learn shielded metal
arc (SMAW or, stick), gas metal arc (GMAW or, MIG), and gas tungsten arc (GTAW or, TIG) welding techniques. The
center employs several welding instructors; two of the instructors participated in this study (Welders 1 and 3). An
additional welder, from the Local 83 Union Hall, also participated in the study (Welder 2).
During a preliminary meeting with the instructors at the Boilermaker's Training Center, it was discussed that
tungsten inert gas (TIG) welding (also known as gas tungsten arc welding) was used primarily for tube work, and metal
inert gas (MIG) welding (also known as gas metal arc welding) was used for buildup work on walls, however, stick
welding was the technique performed most frequently in the field. Therefore, the focus of the study was narrowed to
the evaluation of stick welding of stainless steel. Four standard types of stainless steel electrodes were selected
for evaluation: AWS 308, AWS 309-16, AWS 316, and AWS 347. Four electrode diameters were also selected: 3/32",
1/8", 5/32", and 3/16". Diameters smaller or larger than these four are not commonly used by the
Boilermakers. The electrodes (rods) were from a number of manufacturers, including Alloy Rods (Hanover, PA), Harris
Welco Alloys (Kings Mountain, NC), McKay (Troy, OH), Tech Alloy (Baltimore, MD), and Lincoln (Cleveland, OH).
Twenty-five sample runs of welding were performed overall, with each run lasting 15 minutes. Twenty-three of the
sample runs were conducted outside in a semi-enclosed tank. The tank was located in the rigging area next to the
welding school, and was 12' tall and 20' in diameter (see Figure 1). More than half of the tank's roof was missing
during the study. The tank was constructed of several 6' high plates, with slight gaps between adjoining plates, and
a 6' high, 2' wide opening which served as the entrance. Two workhorses were set up in the tank, with a 9" long
plate of stainless steel affixed to each workhorse. During each sample run, the welders would lay several continuous
beads (welds) along the length of the baseplate. Only flat position welding was performed. Welder 2 was always
positioned on the right side of the tank (when looking through the tank entrance). Welders 1 and 3 were
interchangeably positioned on the left side of the tank.
The first 16 sample runs (1-16) in the tank were conducted to evaluate welding fume emissions using the different rod
types and diameters (4 rod types x 4 diameters). Welders 1 and 2 were each set up to weld at a workhorse during these
runs, and sample data was collected on them simultaneously. No ventilation was used other than natural dilution
ventilation.
Figure 1: Tank Diagram
The next seven sample runs (17-23) were conducted in the tank to evaluate the local exhaust
ventilation (LEV) units. Welder 2 participated in all seven of these runs. Welder 1 welded simultaneously with Welder
2 during one of the seven runs, while Welder 3 welded simultaneously with Welder 2 during another three of the runs.
The LEV units evaluated during these sample runs were selected from four units supplied by Plymovent Canada
(Mississauga, Ontario):
Unit 1: MEF - Mobile, wheeled fume extractor unit with a 6.56' (2 m) flexible arm
Unit 2: BSFM-2101 - Portable fan unit, on a support stand, with a flexible arm
Unit 3: TK-400 - Portable filter with suction hoses and nozzles
Unit 4: MK-800/3 - Mobile mechanical filter unit with a 9.84' (3 m) extraction arm.
Observing the four units on-site led to the conclusion that the MK-800/3 was slightly too large for the applications.
Also, the TK-400 was deemed impractical as a control device in the field as it required a distance of 3" or less
between the welding fume emission source and the exhaust hood to be effective. Since welding of boilers is usually
not stationary work, this type of unit would need to be moved constantly to maintain the 3" capture distance. As
such, the two LEV units selected for evaluation during this study were the MEF and BSFM-2101 models.
Figure 2 depicts the MEF model (Unit 1). This unit's exhaust flexible arm was 160 mm (6.25") in diameter, and
was made of flame proof, double skin, PVC coated woven polyamide with an internal steel spiral. The hood at the end
of the arm was somewhat conically shaped. According to the manufacturer's product literature, the recommended air
flow at the hood of Unit 1 should be between 800-1200 m3/hour (470-706 cfm). With a 10
meter (32') outlet duct attached to the unit, the approximate airflow at the hood is expected to be 1000 m3/hr
(588 cfm). The free flow air volume is designed at 1000 m3/hr (825 cfm). A ½ horsepower
(HP) motor powers the fan. Unit 1 weighed approximately 35 kg (77 lbs), however handles and two front wheels enable
it to be moved with ease.
The BSFM-2101 model (Unit 2) is shown in Figure 3. Unit 2's exhaust arm was similar to that of Unit 1, except that
the hood was not as conically shaped. Product literature indicated that Unit 2 has a 1 HP motor and a free flow air
volume of 1300 cfm.
Neither of the local exhaust ventilation units were equipped with filters during the study. Instead the captured
fumes were exhausted via flex-duct to a point outside the tank.
The final two sample runs of this study (24-25) were conducted inside the welding school with Welder 3. The room
where the welding occurred was approximately 70' long by 34' wide, with about a 10' high ceiling (see Figure 4). The
garage door to the building was kept open about 6" during both runs. Welder 3 sat approximately 6' from the
garage door. Local exhaust ventilation was supplied by Unit 1 during sample run 24, while natural ventilation alone
was used during sample run 25.
Details on all 25 sampling runs can be found in Table 1, including the temperature and humidity data collected at the
start of each sample run.
Conventional industrial hygiene air sampling was performed during the study. Samples were collected in the welders'
breathing zone and in general areas, using closed-faced, 37-millimeter (mm), polyvinyl chloride (PVC)
filters.
Two personal samples were collected simultaneously in the worker's breathing zone using high volume pumps set at a
flow rate of 13 liters per minute (lpm). A length of Tygon® tubing tethered the filters on the welder to
the pumps on the floor. The tubing length allowed the welder to work with minimal restriction during sampling. The
filters were placed on the lapel of the
welders' work shirts, just outside of their welding helmets, since the purpose of the study was to evaluate the
control effectiveness of the ventilation, not the personal protective gear. A distance of about 25" was
maintained between Welder 1's face and the weld arc, while about 20" existed between Welder 2's face and the
arc. Distances were not measured for Welder 3. The filters were replaced with new filters at the beginning of each
run.
Figure 2: Local Exhaust Ventilation Unit 1 (MEF model)
Figure 3: Local Exhaust Ventilation Unit 2 (BSFM model)
Figure 4: Room Layout
Of the two personal samples collected on each welder, one filter was analyzed gravimetrically to determine the total
welding fume concentration. The analysis was conducted according to Method 0500 (for total particulate) in the NIOSH
Manual of Analytical Methods, 4th edition.7 In this method, a known volume of air is drawn
through the pre-weighed PVC filter. The weight gain of the filter is then used to compute the micrograms
(µg) of particulate per cubic meter (m3) of air. After determining the total welding fume
weight on the filter, an element specific analysis was performed on the filter samples, according to NIOSH Method
7300 (modified for microwave digestion). In this method, the different metal species in the welding fume are
differentiated and quantified using an inductively coupled plasma emission spectrometer.
The second personal filter sample collected on the welder was analyzed specifically for hexavalent chromium by
visible spectroscopy, according to NIOSH Method 7600.
In addition, an area sample was collected during all the sample runs using a carbon vane pump set at a rate of 13 lpm.
The area sample was located in the middle of the tank, about
Table 1: Sampling Run Information
Run
Day
#
Welders
Rod
Type
Rod
Diam
Rod Mfg
LEV
Welder 1
LEV
Welder 2
LEV
Welder 3
Temp
°F
Relative
Humidity
1
4/23
2
308
3/32"
Alloy Rods
None
None
-
55
38%
2
4/23
2
309
3/32"
Welco
None
None
-
-
-
3
4/23
2
316
3/32"
McKay
None
None
-
69.2
22%
4
4/23
2
347
3/32"
Welco
None
None
-
70
22%
5
4/23
2
308
1/8"
Alloy Rods
None
None
-
71
22%
6
4/23
2
309
1/8"
Weldco
None
None
-
72.6
20%
7
4/23
2
316
1/8"
Lincoln
None
None
-
73.3
21%
8
4/23
2
347
1/8"
McKay
None
None
-
72.9
21%
9
4/24
2
308
5/32"
McKay
None
None
-
63.3
33%
10
4/24
2
309
5/32"
McKay
None
None
-
62.8
38%
11
4/24
2
316
5/32"
McKay
None
None
-
63.7
37%
12
4/24
2
347
5/32"
McKay
None
None
-
66
37%
13
4/24
2
308
3/16"
McKay
None
None
-
67.9
36%
14
4/24
2
309
3/16"
McKay
None
None
-
70
34%
15
4/24
2
316
3/16"
McKay
None
None
-
81.2
26%
16
4/24
2
347
3/16"
McKay
None
None
-
83
23%
17
4/24
2
308
3/16"
McKay
Unit 2
None
-
84.7
22%
18
4/25
1
308
3/16"
Alloy Rods
-
Unit 2
-
68.4
46%
19
4/25
1
308
3/16"
Alloy Rods
-
Unit 1
-
67.9
39%
20
4/25
1
308
3/16"
Alloy Rods
-
None
-
68.8
34%
21
4/25
2
308
3/16"
Alloy Rods
-
Unit 1
Unit 2 - 9 min
69.4
25%
22
4/25
2
308
3/16"
Alloy Rods
-
Unit 1
Unit 2
71.3
22%
23
4/25
2
347
3/16"
McKay
-
Unit 1
Unit 2
-
-
24
4/25
1
308
3/16"
Tech Alloy
-
-
Unit 1
-
-
25
4/25
1
308
3/16"
Tech Alloy
-
-
None
-
-
*Unit 1: MEF LEV model (small elephant trunk on wheels)
*Unit 2: BSFM LEV model 2101 - Plymovent (fan)
3' off the floor during sample runs 1-23. During sample runs 24-25 the area sample was located 60" from the arc
at a height of 60" off the floor. In all cases, the area sample was within 10' of the welders, thus
approximating personal exposures more than emission source levels. The area samples were analyzed for total welding
fume and elements according to the NIOSH methods listed previously.
Four full term area samples were collected each day of the survey to obtain the background level of air contaminants.
The filters were connected by Tygon® tubing to sampling pumps (SKC Inc., Eighty Four, PA) which ran at a
constant flow rate of 3 lpm. A set of two samples was collected inside the tank while another set of samples was
collected outside the tank. The inside set was located about 3' off the floor, closer to Welder 2 than Welder 1. The
outside set was located several feet away from the entrance to the tank. In each sample set, one filter was analyzed
for total welding fume/elements and the other was analyzed for hexavalent chromium, in the same manner as the
personal samples.
For each of the analyses, there is a limit of detection (LOD) and a limit of quantitation (LOQ). The LOD refers to
the lowest measurable amount on the filter while the LOQ refers to the level at which the laboratory can confidently
report precise results. Appendix D lists the limits of detection and quantification for all the elements analyzed by
the laboratory.
To collect personal sampling data, an aerosol photometer, the Hand-held Aerosol Monitor (HAM) (PPM Inc., Knoxville,
TN), was positioned on one welder's chest using a belt and harness. During the first 17 sample runs, the HAM was worn
by Welder 1. During runs 18-23, Welder 2 wore the HAM, and during runs 24 and 25, Welder 3 wore the HAM. A personal
pump operating at 3 lpm was used to draw air through the HAM's sensing chamber. A filter cassette was mounted on the
HAM to collect the welding fume before it reached the pump. This filter cassette was analyzed for total welding fume
and elements in the same manner as the other filter samples. Only one filter was used per day on the HAM.
Another HAM was used to collect area sampling data. This HAM was positioned on the wall of the tank, close to Welder
2's work area, at a height of 6'. The distance between the area HAM and the welding arc of Welder 2 was noted to be
about 58". For runs 24 and 25, inside the building, the area HAM was located 18" above the floor, 95"
from the center of the baseplate. Due to pump shortages, the area HAM was maintained as a passive sampling device, so
no filter was used.
The HAM operates such that it emits a light from a light-emitting diode. This light is scattered by the aerosol and
forward-scattered light is detected. The amount of scattered light is proportional to the analog output of the HAM.
However, the calibration of the HAM varies with aerosol properties such as refractive index and particle size.8
Therefore, HAM measurements are often expressed as "relative exposures" or "the HAM analog
output", with units of volts. During the first run, the personal HAM was set at a sensitivity level of 20 mg/m3
with a one second averaging time constant. Using this sensitivity level, the analog output of one volt was equated to
a total welding fume concentration of 10 mg/m3 for a calibration dust. After observing the
amount of fume generated during the first run, the sensitivity level was changed to 200 mg/m3
which equates to a total welding fume concentration of 100 mg/m3 per volt. This prevented the personal HAM
from "peaking out" during the data collection. The area HAM was set at a sensitivity level of 2 mg/m3,
equating to a total welding fume concentration of 1 mg/m3 per volt.
The analog output of the HAMs was recorded by Metrosonics data loggers (Model dl-3200, Metrosonics, Inc., Rochester,
NY). When the data collection was completed, the data loggers were downloaded to a personal computer for storage and
analysis. The workers' activities were simultaneously recorded on video for use in a detailed task analysis of the
welding operations.
Optical particle counters (Model 227, Met One, Grants Pass, OR) were also used to obtain information on aerosol
concentrations on the welders and in the general area. When used as a personal sampler, the instrument was clipped
onto one welder's belt and the inlet was positioned in his breathing zone. Welder 1 wore the Met One during the first
17 sample runs. Welder 2 wore the Met One during runs 18-23, and Welder 3 wore the Met One during runs 24-25.
A 30-cm length of 5-mm inside diameter Tygon® tubing was used to transport the
aerosol from the sensor to the instrument.
To monitor total welding fume particle counts in the tank, a second Met One was placed on the tank wall, close to
Welder 1's work area, at a height of 6' 6". A distance of about 65" was noted between the area Met One and
Welder 1's welding arc. After the first 17 runs, the area Met One was moved closer to Welder 2, where it was
positioned next to the area HAM, at a distance of 51" from the center of the baseplate. The Met One remained in
this position for sample runs 18-20 while Welder 2 was the only person welding. Then, when Welder 3 joined Welder 2,
during sample runs 21-23, the Met One was moved back to its original position. For sample runs 24 and 25, inside the
building, the area Met One was located 18" off the floor, at a point 95" from the center of the baseplate
(next to the area HAM).
The Met One instruments continuously record the number of particles counted during a series of consecutive sampling
periods. During this study, a sampling rate of 2.83 lpm, a sampling period of one minute, and a time between sampling
periods of one second were set. Two channels were used to store the number of particle counts in a time interval. One
channel stored the total number of particles counted greater than 0.3 µm. The second channel was set to count the
number of particles larger than 3.0 µm. The particles were sized, based upon the amount of scattered light detected
by the photo detector. In reality, the magnitude of the light pulse scattered by the particles varies with particle
size, optical properties, and surface roughness. The stored data was downloaded directly to the computer to obtain
the particle counts for each sampling period. (According to the instrument manufacturer, the recorded time of the
sample was the end time of the sampling segment.) This data could then be correlated with the videotape to determine
relationships between events and exposures.
The ventilation systems were assessed by measuring capture and face velocities with a hot wire anemometer (TSI
VelociCalc). This instrument measures air velocities in feet-per-minute (fpm) and air volumes in cubic
feet-per-minute (cfm). Capture velocities were measured to determine the ability of the system to remove welding
fumes at certain distances away from the fume generation source. The capture velocity is the velocity necessary to
overcome opposing air currents, thus allowing the welding fume to be exhausted. Face velocities were measured to
compute air volumes. Work methods regarding welding techniques and the use of the ventilation systems were observed.
In addition, airflow patterns around the workers during welding were observed using smoke tubes and aspirators. From
this, an understanding of how air contaminants are transported into the worker's breathing zone can be developed.
Measurements were collected during each sample run for NO2, CO, and O2,
using a PhD Ultra Multi Gas Detector (Biosystems, Inc., Rockfall, CT ). Temperature and relative humidity
measurements were collected using a thermohygrometer.
It should be noted that several study parameters could not be held constant and these factors may have influenced the
data results. The parameters include worker habits, wind/air currents, and temperature and humidity fluctuations.
Observations regarding each of these parameters are discussed below.
The welders work habits were all somewhat different. Welder 1 stood fairly erect with his face positioned directly
over the base plate. Welder 2 also stood, but bent over at the waist when welding, resting his arms on the workhorse,
and keeping his face angled slightly away from the weld area. Welder 3 sat on a stool when welding. Welders 1 and 2
were right handed; Welder 3 was left handed. The welders worked at approximately the same rate, using about the same
number of electrodes during a single run. Throughout the survey, the total number of electrodes used by each welder
during a sample run fluctuated between 10 and 21. This was primarily dependent upon the diameter of the electrode; as
the diameter increased, the number of rods used during the 15 minute run decreased.
The effect of the wind was not well documented. During Runs 3 and 9 the air current appeared to be directed towards
Welder 2, while during Runs 4 and 7 it appeared to be directed towards Welder 1. The second day of sampling appeared
to be slightly windier than the first day. On the third day of sampling, a length of fabric was fastened to a nail on
the tank wall near Worker 2 to serve as a makeshift indicator of the wind direction. During Runs 18 and 19, the wind
was mostly blowing towards the left side of the tank, directly into Welder 2's face. During run 21, the wind appeared
to mostly blow towards the right side of the tank, and during runs 20, 22, and 23 the wind was intermittently blowing
towards either side of the tank. Air velocities inside the tank were generally measured between 20-40 fpm, but
occasional gusts were felt.
As mentioned previously, temperature and relative humidity measurements are shown in Table 1 for each sampling run.
During the first day of sampling, a peak temperature of 73.3°F and a peak humidity level of 38 percent were
measured. Peaks on the second day of sampling were at 84.7°F and 38 percent RH, while on the third day the peaks
were 71.3°F and 46 percent RH. Typically, as the day progressed, the temperature increased and the humidity
decreased.
The LEV units were positioned 3" away from the end of the 9" long base plate. Face velocities were not
measured on Unit 1. However, a face velocity of 1820 fpm was measured at the midpoint of the hood face of Unit 2.
This computes to an airflow of about 390 cfm with the exhaust hoses attached.
Capture velocities were measured for the two ventilated units and are shown in Table 2.
According to Figure VS-90-02 "Welding Ventilation - Movable Exhaust Hoods" in the Industrial Ventilation
Manual, at a distance of up to 6" from the hood, the rate of exhaust should be 250 cfm for a cone-shaped hood or
335 cfm for a plain hood. At distances of 6-9" from the hood, the rate of exhaust for a cone hood should be 560
cfm, or 755 cfm for a plain hood.10 Noting that the hood on Unit 1 was slightly more
conical than the hood on Unit 2,
Table 2: Capture Velocities for the Local Exhaust Ventilation Units
Distance from Hood (in)
Unit 1 (fpm)
Unit 2 (fpm)
6
120
300
9
60
220
12
30
50
the volume of air moved by Unit 1 at a point 6" from the hood was about 325 cfm, and around 800 cfm for Unit 2.
The airflow was approximated using the following equation:
Q = V(10X2 + A)
where:
Q
=
air flow, cfm
V
=
centerline velocity at X distance from the hood, fpm
X
=
distance outward along the axis, ft
A
=
area of hood opening, ft2
The Industrial Ventilation manual also states that the above equation is only accurate for limited
distances of X, where X is within 1.5 times the diameter of the hood. For distances greater than this, the flow rate
increases less rapidly.
Nitrogen dioxide (NO2) and carbon monoxide (CO) were not detected by the sampling
equipment. Oxygen (O2) levels ranged between 20.6 and 21.8 percent. The OSHA PEL for NO2
is 5 ppm (ceiling value), the PEL for CO is 50 ppm, and the O2 level should measure about
21 percent.
The filter data were not converted to 8-hour time weighted averages (TWAs) as the purpose of this study was to
evaluate the welding process and was not compliance driven. The SAS General Linear Models Procedure was used to
perform several analyses of variance (ANOVA) on the filter data to evaluate the effect of study parameters such as
electrode type, electrode diameter and ventilation, on worker exposure. The statistical analyses established a
confounding relationship between the filter data results and both temperature and humidity, suggesting the
temperature and humidity data are correlated with the exposure data.
The results of the personal and area sampling data for total welding fume concentrations can be found in Appendix E.
Out of a total of 45 personal samples, 29 were found to be in excess of the 5 mg/m3 TLV for
welding fumes. Of these 29 samples, fifteen were measured in Welder 1's breathing zone, ten were measured on Welder
2, and four were measured on Welder 3. For the area sampling data, only 1 out of 30 samples was found to be in excess
of the 5 mg/m3 TLV (a 3/16" 308 rod). Figure 5 shows the difference between the welder
data and the area data collected during the 25 sample runs.
Table 3 depicts the various combinations of rod types and diameters used during the study. For each combination, the
table shows the number of personal samples greater than the TLV as compared to the total number of samples collected.
Overall, the 308 and 347 rod types resulted in the greatest percentages of personal over exposures; approximately 70
percent of the personal samples collected with these rods were greater than 5 mg/m3. The
rod diameter rod resulting in the greatest percentage of personal over exposures during sampling was the 3/16"
rod; 71 percent of all the samples collected with this rod were above 5 mg/m3.
Seven out of ten personal samples collected while local exhaust ventilation units were operational (70 percent)
resulted in welding fume levels greater than 5 mg/m3. These samples are included in the
data shown in Table 3. Three (43%) of the over exposures were measured when welding with Unit 1 and four (57%) were
measured when welding with Unit 2. This implied the local exhaust ventilation units were not effectively reducing the
welding fume levels in the worker's breathing zone to below the TLV.
Table 3: Number of Personal Welding Fume Exposures > 5 mg/m3
per Total Number of Samples for each Rod Type and Diameter Combination
Rod Type/Diameter
308
309
316
347
TOTALS
3/32"
1/2
1/2
1/2
2/2
5/8 (63%)
1/8"
2/2
0/2
1/2
1/2
4/8 (50%)
5/32"
2/2
1/2
1/2
1/2
5/8 (63%)
3/16"
8/13
2/2
2/2
3/4
15/21 (71%)
TOTALS
13/19
(68%)
4/8
(50%)
5/8
(63%)
7/10
(70%)
29/45
(64%)
Figure 5: Total Welding Fume Filter Data
To further analyze the impact of ventilation on the total welding fume exposure data, statistical analyses were
performed. From the ANOVA table, ventilation was shown to have a significant effect on the exposure data (p<0.02).
Thus, although the ventilation did not always control the fumes to below the TLV, it did help to reduce the welders' fume exposure levels significantly.
Upon comparing least square means, a significant difference was found between Unit 2 and Unit 1 (p=0.03), and between
Unit 2 and no ventilation (p=0.006). In other words, the welders' exposures when using Unit 2 were significantly
lower than when using Unit 1, or when not using ventilation at all. The ability of Unit 2 to capture the visible
welding fumes is apparent from Photo 1. The welders' exposure data when using Unit 1 was lower than when no
ventilation was used, however the difference was not statistically significant (p=0.64).
A statistical analysis of the data from the first 16 sample runs was performed to identify differences between worker
exposures. A significant difference was found between the total welding fume concentrations measured on Welder 1 and
Welder 2 (p<0.0001), with Welder 2 having significantly lower exposure data. Data for Welder 3 was not analyzed as
he did not participate in the first 16 runs. The humidity was found to significantly affect the measured difference
between Welders 1 and 2 (p<0.001).
Photo 1: Unit 2 Effectively Exhausted the Welding Fumes Away From the Work Area
Since Welder 1 was found to have significantly higher exposures than Welder 2, the previous conclusion, that Unit 1
was not as effective a ventilation control as Unit 2, may be somewhat premature. It could be argued that Unit 2 only
appeared more effective since it was used to control Welder 1's exposures and thus, the exposure reductions would
appear more dramatic than any reductions found for Unit 1 which was not used by Welder 1. If Welder 1 had used Unit
1, this control's fume reduction capabilities may also have been found to be dramatic (significant).
Sample runs 1-16 were also evaluated to determine the effect of rod type and rod diameter on the welders' total fume
concentrations. Statistical analyses found that the rod type and diameter had no significant effect on the total
welding fume exposure data. However, the humidity and temperature data, which were significantly related to the fume
exposure, were confounded by the rod diameter. It is theorized that the relation between temperature and humidity,
and rod diameter, arose due to the order of the sample runs. Rather than be randomized, the sample runs were set up
to facilitate the work of the welders; all four rod types of the same diameter were sampled one after the other, so
that the voltage and current on the welding machines did not need to be continually switched. As such, diameters of
the same size were done in blocks of time, beginning with the smaller diameters. However, as each day of sampling
progressed, the temperature increased and humidity decreased. The lack of randomization may have resulted in the
temperature and humidity data appearing correlated with the increasing diameter size.
The elemental analysis showed that out of 28 measured elements, 5 metals were of concern: arsenic, chromium, iron,
manganese, and nickel (see Appendix F). Forty-four of the 45 personal samples (98 percent) and eight of the 30 area
samples (27 percent), were found to have at least one of these five fume constituents above the applicable elemental
exposure guidelines. Manganese and nickel alone accounted for 70 percent of the personal and area over exposures (see
Figure 6).
The REL of 15 µg/m3 is the most stringent guide for nickel exposures. The highest nickel
concentration measured was 667 µg/m3 (45 times the REL). A TLV of 200 µg/m3
is the most stringent manganese guide. The highest concentration of manganese measured was 3692 µg/m3
(19 times the TLV). The REL of 2 µg/m3 is the most stringent arsenic guide. The highest
concentration of arsenic measured was 16 µg/m3 (8 times the REL). The REL and TLV are the
most stringent guides for total chromium and iron (as iron oxide); the REL/TLV for total chromium is 500 µg/m3
and the REL/TLV for iron is 5000 µg/m3. The highest total chromium exposure measured was
3846 µg/m3 (8 times the limits) and the highest iron exposure measured was 5128 µg/m3
(just barely over the limits). Of the five metals, only arsenic (seven samples) and chromium (ten samples) were
found in concentrations above the OSHA PEL levels. Also, upon reviewing the data, the rod types and diameters did not
appear to greatly influence the exposure levels.
Figure 6: Five Elements Account for 100% of the Overexposures
The over exposures discussed above included ten personal and six area samples collected while the local exhaust
ventilation units were operational. Out of these 16 samples, the number of over exposures for each of the five metals
were as follows: 2 arsenic over exposures (13 percent), 3 chromium (19 percent), 0 iron, 8 manganese (50 percent),
and 13 nickel (81 percent). Both arsenic overexposures exceeded the OSHA PEL; the other metal overexposures did not
exceed PELs, but were greater than the RELs or TLVs. Again, the local exhaust ventilation did not appear to be
effectively controlling the welding fume constituents to below the recommended exposure levels.
The results of the personal and area sampling data for hexavalent chromium fume concentrations can be found in
Appendix G. The majority of samples were greatly in excess of the 50 µg/m3 ACGIH TLV and
the 1 µg/m3 NIOSH REL for hexavalent chromium. (NIOSH considers hexavalent chromium to be
a potential occupational carcinogen.) It should be noted that no OSHA PEL exists specifically for hexavalent
chromium, however OSHA does enforce a ceiling value of 0.1 mg/m3 for chromic acid (CrO3)
and chromates.11 In addition, OSHA is currently responding to a petition by Public Citizen
and the Oil, Chemical and Atomic Workers for an emergency standard to reduce worker exposures to hexavalent chromium.12
High exposures occurred regardless of whether or not ventilation was being used during the welding operations. Nine
personal samples had exposures greater than 1000 µg/m3; eight of these were on Welder 1
and one was on Welder 3. The 3/16" and 5/32" diameters of each rod type (308, 309, 316, and 347) resulted
in the exposures for Welder 1. The 3/16" diameter 308 rod resulted in the exposure to Welder 3, with Unit 2
operational.
Differences in personal concentrations between the welders are shown in Figure 7. The highest concentration of
hexavalent chromium (2716 µg/m3) was collected on a filter that was attached to the HAM on
Welder 1. This filter remained on the HAM during sample runs 9-17. The only area samples collected for hexavalent
chromium were full-term samples. No area exposures were found to occur when sampling outside the tank; however the
full-term area exposure levels when sampling inside the tank ranged from 4-19 µg/m3.
Statistical analyses of the hexavalent chromium data collected during the first 16 sample runs found the same results
as the total welding fume analysis. A statistically significant difference (p<0.0001) existed between the
concentrations measured on Welders 1 and 2; this difference was affected by the humidity data (p<0.0001). The type
of welding rod and the diameter of the rod, again, had no direct effect on the concentration data.
Figure 7: Hexavalent Chromium Filter Data
Additional statistical analyses were performed to further evaluate the effect of ventilation on the hexavalent
chromium fume exposure data. The use of ventilation was shown to have a significant effect on reducing the exposure
data (p<0.04), even though all the samples collected with ventilation were still above the TLV. Upon comparing
least square means, a significant difference was found to exist between Unit 2 and no ventilation (p=0.02). In other
words, the use of Unit 2 significantly reduced the amount of hexavalent chromium fume in the welders' breathing zone
when compared to welding with no ventilation. The exposure to hexavalent chromium fume when using Unit 1 was not
statistically significantly different from when no ventilation was used (p=0.9). Overall, the hexavalent chromium
fume levels were lower when the welders used Unit 2 as compared to Unit 1, however this difference was not proven to
be statistically significant (p=0.09).
The personal and area average relative concentration data, as measured by the HAM, appeared to be lowest when the
ventilation was operational (Figure 8). The highest average area relative concentration occurred when the welding
operation was located inside the building with no ventilation (sample run 25). No area data was retrieved for sample
run 19. The highest average personal concentrations occurred during sample runs 9, 11, and 14. These three runs all
occurred while the HAM was located in the breathing zone of Welder 1. A comparison of the HAM
results (Figure 8) and the total welding fume filter data results (Figure 5), shows the two data sets follow similar
patterns over the 25 sampling runs. However the relative concentrations as measured by the HAM in terms of mg/m3
were much higher than what was actually measured using the filters.
Figure 8: Total Welding Fume Relative Concentration Data as
Collected by the HAM
The average particle count data for the personal and area samples, as collected by the Met Ones, are summarized in
Figures 9 and 10. No area data was retrieved for sample run 16. For particles greater than 0.3µ, the average count
measured during a sample run was often lower in the welder's breathing zone than the count measured by the area
monitor (Figure 9). This may have been due to the effect of the air currents on the welding fume. Also, during the
first 16 sample runs, higher average particle counts were measured when welding with rods in the smaller two
diameters than in the larger two diameters. This was more noticeable for particles > 0.3µ than for particles >
3.0µ.
Figure 9: Particle Count Data (> 0.3u) as Collected by the
Met One
Figure 10: Particle Count Data (> 3u) as Collected by the Met
One
The data collected during sample runs 24 and 25 inside the building showed that the local exhaust ventilation helped
to reduce fume exposures to the worker. This was most evident when analyzing the filter data; the total welding fume
exposures for the personal and area samples were both 5 times lower with the ventilation on than with it off (see
Figure 5). The hexavalent chromium filter data collected on the worker also showed the ventilation to reduce fume
exposure by a factor of five (Figure 6). The personal HAM data showed the relative concentration dropped by a factor
of 4 with ventilation, while the area HAM data was 3 times lower with the ventilation (Figure 8). The Met One
personal and area particle count data were also lower when using ventilation; these differences are more clearly
shown in Figures 11 and 12. The reduction found for larger particles, greater than 3 microns in diameter, is depicted
in Figure 11, while the reduction found for all particles greater than 0.3 microns in diameter is depicted in Figure
12.
The results of the sampling data showed that the welders were exposed to high levels of stainless steel welding fume
almost two thirds of the time during the study. Several of the personal samples were extremely high; the highest
level of welding fume measured on a welder was 60 mg/m3, almost 12 times ACGIH's TLV of 5
mg/m3. In addition to exceeding the recommended levels for total welding fume, many of the
personal and area samples exceeded levels set for arsenic, total chromium, hexavalent chromium, iron, manganese, and
nickel. Nickel and hexavalent chromium are both considered potential occupational carcinogens by NIOSH. Personal
exposure levels reached almost 0.7 mg/m3 for nickel (REL=0.015 mg/m3),
and almost 3 mg/m3 for hexavalent chromium (REL=0.001 mg/m3).
A significant difference was established between the fume concentration measured on Welder 1 and Welder 2 using the
filter data. Welder 1 was found to have significantly higher levels of welding fume concentrations than Welder 2. It
is unclear how much of the difference between the two welders' exposures can be attributed to wind direction and to
work methods (Welder 1 stood erect but had his face in a direct line with the welding plume, while Welder 2 leaned
over his work but kept his face at an angle from the plume). Air currents in the tank, on average, ranged between 20
and 40 fpm. Statistics found a correlation between the temperature and humidity data and the difference between
worker exposure levels, but this was probably due to the inability to randomize the sampling runs. Because of the
confounding temperature and humidity data, no discernable difference could be established between the fume
concentrations generated by different rod types and diameters.
Measurements collected on the ventilation units showed that Unit 2 (the fan), used by all three welders, had better
capture velocities than Unit 1 (the elephant trunk), which was only used by
Welders 2 and 3. At 6" from the hood, Unit 2 was moving air at about 300 fpm, while Unit 1 was moving air at
about 120 fpm. At 12" from the hood (the furthest point the welders worked), Unit 2 was moving 50 fpm, while
Unit 1 moved about 30 fpm.
Figure 11: Effect of Ventilation on Particle Count Data
(> 3u) Inside the Building
Figure 12: Effect of Ventilation on Particle Count Data
(> 0.3u) Inside the Building
Statistics showed there was a difference in the effectiveness of the two types of local exhaust ventilation units
evaluated. Unit 2 was found to be better at removing the welding fumes than Unit 1, and was significantly better than
when no ventilation was used. And, although using Unit 1 did help to reduce welding fume levels, it was not found to
be significantly better than through natural ventilation methods.
In summary, stainless steel welding can result in high fume exposures to workers. When the workers are welding
outside, even in a semi-enclosed tank, air currents play a significant role in how much fume is carried
into the worker's breathing zone. If the wind is directed towards the welder, his fume exposure is going to increase.
Even the use of local exhaust ventilation may not have a significant effect on reducing the worker's exposure. The
worker's exposure level is going to be dependent on how strong the wind is and in what direction the wind is moving,
where the worker is standing in relation to the welding plume, and where the ventilation is positioned. As such, an
adequate reduction in worker exposures by using local exhaust ventilation when welding outside is not guaranteed,
even in a semi-enclosed area, due to the potentially strong effects of even a slight wind current. Ventilation will
help to reduce fume exposures but the ability of the welder to always stand upwind of the fumes may even be more
important when working outside.
1. AWS Safety and Health Committee [1993]. Effects of welding on health VIII. Miami, FL: American Welding Society,
Safety and Health Committee; ISBN: 0-87171-437-X.
2. Hewett P [1988]. Summary ranking of small businesses most in need of control technology from 23 OSHA state
consultation programs. Memorandum of December 20, 1988, from P. Hewett, Division of Respiratory Disease Studies, to
Small Business Initiative Committee, National Institute for Occupational Safety and Health, Centers for Disease
Control and Prevention, Public Health Service, U.S. Department of Health and Human Services.
3. Webster HK [1995]. Welding could become an OSHA priority. Welding Journal 2:7.
4. NIOSH [1988]. Criteria for a recommended standard: occupational exposure to welding, brazing, and thermal cutting.
Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control,
National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 88-110.
5. NIOSH [1994]. NIOSH pocket guide to chemical hazards. Cincinnati, OH: U.S. Department of Health and Human
Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational
Safety and Health, DHHS (NIOSH) Publication No. 94-116.
6. ACGIH [1994]. Threshold limit values for chemical substances and physical agents and biological exposure indices.
Cincinnati, OH: American Conference of Governmental Industrial Hygienists.
7. NIOSH [1984]. Nuisance dust total, Method 0500. In: NIOSH Manual of Analytical Methods, 3rd ed., with supplements
1, 2, 3, and 4. Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for
Disease Control, National Institute for Occupational Safety and Health, NIOSH Publication No. 84-100.
8. NIOSH [1992]. Analyzing workplace exposures using direct reading instruments and video exposure monitoring
techniques. Cincinnati OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease
Control and Prevention, National Institute for Occupational Safety and Health DHHS (NIOSH) Publication No. 92-104.
9. Gressel MG, Heitbrink WA, McGlothlin JD, Fischbach TJ [1987]. Advantages of real-time data acquisition for
exposure assessment. Appl Ind Hyg 3(11):316-320.
10. ACGIH [1995]. Industrial ventilation, a manual of recommended practices. 22nd ed. Cincinnati, OH: American
Conference of Governmental Industrial Hygienists.
11. CFR. Code of Federal regulations [1994]. Occupational Safety and Health Administration: OSHA Table Z-2. 29CFR
1910.1000. Washington, D.C.: U.S. Government Printing Office, Federal Register.
12. Bureau of National Affairs [1996]. BNA Occupational Safety & Health Daily, Article No. 32741404. Washington,
D.C. September 30, 1996.
The National Institute for Occupational Safety and Health (NIOSH) is located in the Centers for Disease Control and
Prevention (CDC), under the Department of Health and Human Services (DHHS) (formerly the Department of Health,
Education, and Welfare). NIOSH was established in 1970 by the Occupational Safety and Health Act, at the same time
that the Occupational Safety and Health Administration (OSHA) was established in the Department of Labor (DOL). The
OSHAct legislation mandated NIOSH to conduct research and education programs separate from the standard and
enforcement functions conducted by OSHA. An important area of NIOSH research deals with methods for controlling
occupational exposure to potential chemicals and physical hazards.
The Engineering Control Technology Branch (ECTB) of the Division of Physical Sciences and Engineering (DPSE) has been
given the lead within NIOSH to study and develop engineering controls and assess their impact on reducing
occupational illness. Since 1976, ECTB has conducted a large number of studies to evaluate engineering control
technology based upon industry, process, or control technique. The objective of each of these studies has been to
evaluate and document control techniques and to determine the effectiveness of the control techniques in reducing
potential health hazards in an industry or for a specific process.
During the past twenty years, the National Institute for Occupational Safety and Health (NIOSH) has documented and
reported on the need to control worker exposures to the fumes and gases generated during welding operations. Much of
the attention to welding has been in the form of Health Hazard Evaluations conducted at field sites; however, a few
NIOSH reports have focused on control technology. These reports are briefly discussed below and can be obtained
through NTIS or the NIOSH Publications Office (1-800-35-NIOSH).
In 1974, a research contract report entitled "Engineering Control of Welding Fumes" was published, with
the objective of developing design criteria for local ventilation systems to control welding fumes. This report
identified shielded manual metal arc welding on carbon and stainless steel, and gas-shielded arc welding on carbon
steel as processes constituting great health risks to welders. A crossdraft table, free-standing hood, and low
volume-high velocity fume extraction gun were evaluated to determine the minimum system operating point needed to
reduce fumes below threshold limit values (TLVs).1
In 1978, the NIOSH booklet "Safety and Health in Arc Welding and Gas Welding and Cutting" included
general information on dilution and local exhaust ventilation.2
In 1979, NIOSH's Division of Physical Sciences and Engineering (DPSE) published the research report
"Assessment of Selected Control Technology Techniques for Welding Fumes." This study considered the
effect of dilution airflow direction on welder exposures in the field and evaluated a fume extraction gun.3
In 1988, the NIOSH "Criteria for a Recommended Standard for Welding, Brazing, and Thermal Cutting" was
produced. In this document, NIOSH recommended that welding emissions be controlled to concentrations as low as
feasibly possible using state-of-the-art engineering technology and work practices. General guidelines
were provided for selecting dilution and local exhaust ventilation systems.4
REFERENCES
1. Astleford W [1974]. Engineering control of welding fumes. Southwest Research Institute, San Antonio, TX. Contract
No. HSM 99-72-76. U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease
Control, National Institute for Occupational Safety and Health, DHEW (NIOSH) Publication No. 75-115.
2. NIOSH [1978]. Safety and health in arc welding and gas welding and cutting. Cincinnati, OH: U.S. Department of
Health, Education, and Welfare, Public Health Service, Center for Disease Control, National Institute for
Occupational Safety and Health, DHEW (NIOSH) Publication No. 78-138.
3. Van Wagenen HD [1979]. Assessment of selected control technology techniques for welding fumes. Cincinnati, OH:
U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control, National
Institute for Occupational Safety and Health, DHEW (NIOSH) Publication No. 79-125.
4. NIOSH [1988]. Criteria for a recommended standard: occupational exposure to welding, brazing, and thermal cutting.
Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control,
National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 88-110.
Welding fumes are a product of the base metal being welded, the welding process and parameters (such as voltage and
amperage), the composition of the consumable welding electrode or wire, the shielding gas, and any surface coatings
or contaminants on the base metal. It has been suggested that as much as 95 percent of the welding fume actually
originates from the melting of the electrode or wire consumable.1 The size of welding fume
is highly variable and ranges from less than 1-µm diameter (not visible) to 50-µm diam (seen as smoke).2
Fume constituents may include minerals such as silica and fluorides (used as fluxes) and metals such as: arsenic,
beryllium (in high copper alloys), cadmium (often used as a rust inhibitor), chromium, cobalt, and nickel (in
stainless steel), copper (in copper-coated wire), iron, lead (in lead-based paint coatings), magnesium, manganese (in
stainless steel, manganese steel), molybdenum, tin, vanadium, and zinc (used to galvanize steel).3,4,5
Toxic gases such as ozone, carbon monoxide, nitrogen dioxide, and phosgene (formed from chlorinated solvent
decomposition) can also be produced.3,4,5 Volatile hydrocarbons can be produced during
welding if antispatter sprays, oils, or lanolin (often used during degreasing processes) are present.2
REFERENCES
1. Stern RM [1979]. Control technology for improvement of welding hygiene, some preliminary considerations.
Copenhagen, Denmark: The Danish Welding Institute, The Working Environment Research Group, ISBN 87-87806-18-5, p. 2.
2. The Welding Institute [1976]. The facts about fume - a welding engineer's handbook. Abington, Cambridge, England:
The Welding Institute.
3. NIOSH [1988]. Criteria for a recommended standard: occupational exposure to welding, brazing, and thermal cutting.
Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control,
National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 88-110.
4. American Welding Society [1987]. Welding handbook. 8th ed., Vol. 1, Welding technology. Connor LP, ed. Miami, FL:
American Welding Society, ISBN: 0-87171-281-4.
5. Rekus JF [1990]. Health hazards in welding. Body Shop Business 11:66-77, 188.