Measuring exposed magnetic fields of welders in working time.

Abstract

The assessment of the occupational electromagnetic field exposure of welders is of great importance, especially in shielded-arc welding, which uses relatively high electric currents of up to several hundred amperes. In the present study, we measured the magnetic field exposure level of welders in the course of working. A 3-axis Hall magnetometer was attached to a subject's wrist in order to place the sensor probe at the closest position to the magnetic source (a cable from the current source). Data was acquired every 5 s from the beginning of the work time. The maximum exposed field was 0.35-3.35 mT (Mean ± SD: 1.55 ± 0.93 mT, N=17) and the average value per day was 0.04-0.12 mT (Mean ± SD: 0.07 ± 0.02 mT, N=17). We also conducted a finite element method-based analysis of human hand tissue for the electromagnetic field dosimetry. In addition, the magnetic field associated with grinders, an air hammer, and a drill using electromagnetic anchorage were measured; however, the magnetic fields were much lower than those generated in the welding process. These results agreed well with the results of the electromagnetic field dosimetry (1.49 mT at the wrist position), and the calculated eddy current (4.28 mA/m(2)) was much lower than the well-known guideline thresholds for electrical nerve or muscular stimulation.

Full text

Page 1

274
Industrial Health 2011, 49, 274–279
S YAMAGUCHI-SEKINO et al.
Original Article
Industrial Health 2011, 49, 274–279
Introduction
Approximately 220,000 people are engaged in weld-
ing operations in Japan. However, the possible adverse
effects of electromagnetic fields (EMFs) generated in
welding processes are of great importance because
welding uses relatively high electric currents of up to
several hundred amperes1–3).
The coupling of low-frequency magnetic fields and
the absorption of energy from EMF is a basic mecha-
nism through which time-varying electric and magnetic
fields interact directly with living matter4–7). In the
physical interaction of a time-varying magnetic field
with a human body, electric fields and circulating elec-
tric currents are induced inside the body. The magni-
tudes of the induced field strength and current density
are proportional to the radius of the loop, the electrical
conductivity of the tissue, and the rate of change in the
magnitude of the magnetic flux density7, 8). These cur-
rents can stimulate nerves. Exposure to an EMF at a
frequency greater than ~100 kHz can cause a consider-
able absorption of energy and a temperature increase.
Because the biological effects of EMF depend on the
field properties, characterization of the frequencies and
the magnetic field strength is essential. The field prop-
erties in welding depend on the process (e.g., shielded-
arc welding, tungsten inert gas (TIG) welding, metal
inert gas (MIG) welding, etc.) and the capability of the
power source (DC, sinusoidal AC, rectangular wave,
etc.)1, 3). In typical welds, low-frequency components
that are lower than 1 kHz are dominant, although some-
Measuring Exposed Magnetic Fields of Welders
in Working Time
Sachiko YAMAGUCHI-SEKINO1*, Jun OJIMA1, Masaki SEKINO2,
Minoru HOJO3, Hiroyuki SAITO1 and Tsutomu OKUNO1
1National Institute of Occupational Safety and Health, 6–21–1 Nagao, Tama-ku, Kawasaki 214-8585, Japan
2Department of Electrical Engineering and Information Systems, School of Engineering, The University of
Tokyo, Chiba, Japan
3Ota Regional Occupational Health Center, Tokyo, Japan
Received December 20, 2010 and accepted April 11, 2011
Abstract: The assessment of the occupational electromagnetic field exposure of welders is of
great importance, especially in shielded-arc welding, which uses relatively high electric currents
of up to several hundred amperes. In the present study, we measured the magnetic field expo-
sure level of welders in the course of working. A 3-axis Hall magnetometer was attached to a
subject’s wrist in order to place the sensor probe at the closest position to the magnetic source
(a cable from the current source). Data was acquired every 5 s from the beginning of the
work time. The maximum exposed field was 0.35–3.35 mT (Mean ± SD: 1.55 ± 0.93 mT, N=17)
and the average value per day was 0.04–0.12 mT (Mean ± SD: 0.07 ± 0.02 mT, N=17). We also
conducted a finite element method-based analysis of human hand tissue for the electromagnetic
field dosimetry. In addition, the magnetic field associated with grinders, an air hammer, and a
drill using electromagnetic anchorage were measured; however, the magnetic fields were much
lower than those generated in the welding process. These results agreed well with the results
of the electromagnetic field dosimetry (1.49 mT at the wrist position), and the calculated eddy
current (4.28 mA/m2) was much lower than the well-known guideline thresholds for electrical
nerve or muscular stimulation.
Key words: Welding, Magnetic fields, Electromagnetic field dosimetry, Eddy current, Welding current
*To whom correspondence should be addressed.
E-mail: yamaguchi@h.jniosh.go.jp

Page 2

MAGNEITC FIELDS IN WELDING PROCESS
275
times high frequencies between several kilo to mega Hz
exist due to the ripple currents generated by the welding
power source1).
So far, a number of reports and guidelines on EMFs
have been published7, 9). The exposure guidelines pub-
lished by the International Commission on Non-Ionizing
Radiation Protection (ICNIRP)7) and the threshold
limit values (TLVs) of the American Conference of
Governmental Industrial Hygienists (ACGIH)9) are wide-
ly accepted in Japan. There have been several reports
containing the measurements of magnetic fields for arc
welding1, 10). A review of the literature indicates that
the ICNIRP reference levels for magnetic fields may
be exceeded in both arc and resistance welding1, 3).
One report shows that for most arc welding processes,
the ICNIRP reference levels are not exceeded unless
the measurement is taken close to the welding cable,
i.e., 10 cm, but for manual welding the reference level
would be exceeded at the welder’s hand1).
Considering the results of this previous report1), the
present study focused on the local exposure to the mag-
netic field at a welder’s hand to investigate the actual
exposure level at the position closest to the magnetic
source, which in this study was the cable from the cur-
rent source. Electromagnetic field dosimetry was per-
formed to compare the results of the magnetic fields
in the workplace and to calculate the eddy currents
induced in a subject’s wrist by the welding processes.
Material and Methods
Measurement of welding currents
The welding currents were obtained by a current
probe (PR430: LEM, Measuring range ± 60 A and ± 600 A,
frequency range: DC to 20 kHz) and a data logger
(8807: HIOKI, sampling rate: 8–400 kHz). Fast Fourier
Transform (FFT) analysis was performed to show fre-
quency components of welding current.
Electromagnetic field dosimetry
We used a finite element method (FEM) –based
analysis for the electromagnetic field dosimetry11). A
two-dimensional human wrist model consisting of
five tissues with different conductivities was devel-
oped. The conductivities of the tissues were obtained
from the database of electric properties developed by
Gabriel et al.12, 13) (skin dry: 2.000 × 10–4 S/m, muscle:
3.001 × 10–1 S/m, bone: 2.008 × 10–2 S/m, blood vessel:
3.005 × 10–1 S/m, and nerve: 2.829 × 10–2 S/m). It was
assumed that the relative magnetic permeabilities of all
the tissues were 1.0. The model was constructed with
a resolution of 2 mm and the number of elements was
37,500. The wire was placed at 6–20 mm from the
wrist model. The results were obtained using a com-
puter program (PHOTO-Series: PHOTON Co., Ltd.).
The electric current applied to the wire had an intensity
of 15 A, 300 Hz, which resulted in a magnetic field
intensity of 2.71 mT at the wire element.
Measurement of magnetic field at the workplace
To measure the actual workplace, 22 subjects (welders:
17 people and office workers: 5 people) were selected
from five industrial companies and two research insti-
tutes in Japan. Magnetic field data was acquired every
5 s by the 3-axis Hall magnetometer (THM1176:
Metrolab, measuring range: 8.0 mT, frequency range:
DC~1 kHz) at a 2.048 kHz sampling rate. The sensor
probe was attached to the subject’s right wrist (Fig. 1)
and measurements were taken throughout the work-
ing time (approximately 8 h except for lunch time).
Welders (17 people) were mainly engaged in welding
operation, however, other work contents (cutting, polish-
ing, hole-drilling, and crane operation) were also includ-
ed. These electromagnetic field sources (drill, grinder
and air hammer) were also recorded individually using
the same protocol. Office workers were mainly engaged
in clerical works using personal computer operation (PC
operation). The background magnetic field (MF) level
Fig. 1. Measurement of magnetic field at the workplace.
Magnetic field data was acquired by the 3-axis Hall magnetometer and
the sensor probe was attached to the subject’s right wrist.

Page 3

276 S YAMAGUCHI-SEKINO et al.
Industrial Health 2011, 49, 274–279
was measured for 5 min before starting operation at
each working place and data was averaged.
Results
Electromagnetic field dosimetry
Figure 2 shows the results of the electromagnetic
field dosimetry. The maximum magnetic flux density in
the model was 1.49 mT. The maximum induced eddy
currents were 4.28 mA/m2 and each value was observed
in the muscle in the model (Table 1).
Measurement of magnetic field at the workplace
Each subject’s belongings, work contents, maximum
and average exposed magnetic fields, and other elec-
tromagnetic field (EMF) sources (drill, grinder and air
hammer) are shown in Table 2. The maximum exposed
field throughout the working time varied between
0.35–3.35 mT (Mean ± SD: 1.55 ± 0.93 mT, N=17),
and the 1-day average exposed field was 0.04–0.12 mT
(Mean ± SD: 0.07 ± 0.02 mT). The average of back
ground level was 0.05 mT. Welders were exposed high-
er magnetic fields than those of office workers (Maximum
exposed field: 0.07–0.09 mT, Table 2). Since welders
were also engaged in other work contents (cutting, pol-
ishing, hole-drilling, and crane operation), these other
electromagnetic field sources (drill, grinder and air ham-
mer) were recorded individually to compare the strength
of MF to that of welding. However, these magnetic
fields were much lower than those of the welding pro-
cess (Table 3). Figure 3 shows a typical welding cur-
rent observed in the current study. Time-varying (AC)
component with static (DC) bias and several irregular
peaks were observed. FFT analysis of typical weld-
ing current showed an obvious peak at DC and a small
peak at 300 Hz due to DC bias in the original source
(Fig. 4A). Then DC filtration was performed in order
to evaluate AC components of the source. Visible peak
was observed at 300 Hz and harmonics were shown up
to 2 kHz in DC-filtered source (Fig. 4B).
Discussion
In the present study, we performed the measurement
of the local MF exposure at welders hand position to
investigate the actual exposure level at the position clos-
est to the magnetic source, which in this study was the
cable from the current source. We used 3-axis Hall
magnetometer which has a acquisition capability in
DC-1 KHz because low-frequency components that are
lower than 1 kHz are dominant in the typical welds.
Relatively high transient MF exposure was observed
throughout the working time of welders (0.35–3.35 mT,
Mean ± SD: 1.55 ± 0.93 mT, N=17), however, the one-
day average (0.04–0.12 mT, Mean ± SD: 0.07 ± 0.02 mT,
N=17) was mostly same to the background level
(0.05 mT).
Since the transient high (-mT) MF exposure may
induce the electrical stimulation in low frequencies
(<100 kHz), electromagnetic field dosimetry was per-
formed to calculate the magnetically-induced eddy
current inside the tissue. The electrical condition was
Fig. 2. Electromagnetic field dosimetry.
A: Numerical model with 2 mm section. B: Induced magnetic field den-
sity in a model. C: Induced eddy current in a model.
Table 1. Electromagnetic field dosimetry by FEM analysis showing
the maximum induced eddy currents in the model
Maximum induced eddy current [mA/m2]
Distance from the wire Whole model Tissue name
6 mm
10 mm
20 mm
4.28
3.80
3.17
Muscle
Muscle
Muscle

Page 4

MAGNEITC FIELDS IN WELDING PROCESS
277
set as 300 Hz, 15A based on the FFT analysis of the
welding current (Figs. 3 and 4). The results (1.49 mT
at the wrist position, Fig. 2) agreed well with the results
of those of the workplace (Table 2), and the calculated
eddy current (4.28 mA/m2, Table 1).
The thresholds of neuronal stimulation lower than
100 mA/m2 were derived from other studies of the visu-
al and mental functions in human subjects. Changes in
the response latency for complex reasoning tests were
reported in individuals subjected to weak power-frequen-
cy electric currents passed through electrodes attached to
the head and shoulders, for which the current densities
were estimated to be between 10 and 40 mA/m2 14, 15).
Also, many studies have reported that subjects expe-
rienced faint flickering visual sensations, known as
magnetic phosphenes, during exposure to extremely low
frequency (ELF) magnetic fields above 3–5 mT16). At
20 Hz, current densities of approximately 10 mA/m2 in
the retina have been estimated as the threshold for the
induction of phosphenes; this level is above the typical
endogenous current densities in electrically excitable
tissues. Higher thresholds have been observed for both
lower and higher frequencies17, 18). However, the maxi-
mum induced eddy currents in the present study were
shown to be 4.28 mA/m2 (4.28 × 10–4 mA/cm2) (Table 1).
Therefore, these results indicate that magnetically
induced eddy currents in the present study are lower
Table 2. Measurement of magnetic field at the workplace
Subject
No
Company Work contentsMax exposed field (mT)*
Average exposed field
(mT)**
Other EMF sources***
1A Shielded-arc welding2.790.05 Grinder
Grinder
Drill#
Grinder
Grinder
Drill#
Grinder
Grinder
Drill#
Grinder
Grinder
Air hammer
Grinder
Drill#
Grinder
Drill#
Grinder
Drill#
Grinder
Drill#
Grinder
Grinder
Grinder
Grinder
PC
PC
PC
PC
PC
-
2AShielded-arc welding3.350.08
3AShielded-arc welding 1.440.05
4A Shielded-arc welding1.67 0.07
5AShielded-arc welding0.990.07
6AShielded-arc welding2.420.07
7
8
9
B
B
C
TIG welding / Rectangular wave
TIG welding / Rectangular wave
Shielded-arc welding
0.6
0.35
0.59
0.04
0.05
0.12
10DShielded-arc welding 0.810.07
11DShielded-arc welding1.000.07
12DShielded-arc welding0.850.06
13DShielded-arc welding 3.050.08
14
15
16
17
18
19
20
21
22
23
E
E
E
E
F
F
F
F
F
-
Shielded-arc welding
Shielded-arc welding
Shielded-arc welding
Shielded-arc welding
PC operation##
PC operation##
PC operation##
PC operation##
PC operation##
Background level
1.70
2.18
0.74
1.77
0.07
0.09
0.07
0.09
0.08
0.06
0.10
0.05
0.05
0.07
0.06
0.06
0.06
0.05
0.06
0.05
*The highest value of exposed field throughout the measurement, **Average of exposed fields throughout the measurement, ***EMF source except
for welding process, #A drill using electromagnetic anchorage, ##Personal computer (PC) operation.
Table 3. EMF sources in the work place
Max exposed field (mT)
Welding
Grinder
Drill*
Air hammer
8.69
1.01
1.70
1.68
*A drill using electromagnetic anchorage.

Page 5

278 S YAMAGUCHI-SEKINO et al.
Industrial Health 2011, 49, 274–279
than the guideline thresholds for neuronal and muscular
stimulation. And this value did not exceed the basic
restriction of ICNIRP guideline (300 Hz, 10 mA/m2).
Cardiovascular systems were also considered to be
sensitive to electric stimulation. However, experimen-
tal studies of both short-term and long-term exposure
indicate that, while electric shock is an obvious health
hazard, other hazardous cardiovascular effects associated
with low frequency fields are unlikely to occur at expo-
sure levels commonly encountered environmentally or
occupationally4).
Evaluation of the EMF exposure level based on fre-
quency components was essential since the exposure
guidelines published by ICNIRP7) or ACGIH9) give the
reference levels or TLVs for the magnetic field accord-
ing to the frequencies. As both ICNIRP and ACGIH
provide separate guidelines for static (DC) and time-
varying (AC) magnetic fields, the DC and AC compo-
nents of the data must be assessed individually if the
acquired data contains the overlapping DC/AC signal
(Fig. 3A). The maximum magnetic field of the DC/
AC superimposed signal was 3.35 mT (Table 1) and
this value was lower than the widely accepted guide-
lines for the static (not time-varying) magnetic field:
ICNIRP guidelines (8 T, occupational)19) and ACGIH
(20 T (ceiling value), limbs)20). Further analysis using
exposure level tester (e.g., Narda ELT-400) is necessary
to evaluate assess the compatibility evaluation for the
ICNIRP and ACGIH exposure guidelines of AC compo-
nent.
Conclusion
The present study focused on the local exposed mag-
netic field at a welder’s hand to investigate the actual
exposure level at the position closest to the magnetic
source, which was a cable from the current source. The
maximum exposed field was 0.35–3.35 mT (Mean ± SD:
Fig. 3. Typical waveforms of welding currents of the present study.
A: Welding current of shielded-arc welding (sampling rate: 8 kHz).
B: Welding current of shielded-arc welding (sampling rate: 400 kHz). AC
component with DC bias and several irregular peaks were observed.
Fig. 4. Fast Fourier Transform (FFT) analysis of the welding cur-
rent.
A: FFT analysis of the original source. B: FFT analysis of DC-filtered
source. An obvious peak at DC and a small peak at 300 Hz were shown
in the original source (Fig. 4A). Visible peak was observed at 300 Hz and
harmonics were shown up to 2 kHz in DC-filtered source (Fig. 4B).