Annals of Occupational Hygiene Advance Access originally published online on August 26, 2005
Annals of Occupational Hygiene 2005 49(8):673-682; doi:10.1093/annhyg/mei031
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2005 British Occupational Hygiene Society Published by Oxford University Press
Original Article |
Equipment Grounding Affects Contact Current Exposure: A Case Study of Sewing Machines
1 Enertech Consultants, 300 Orchard City Drive, Suite 132, Campbell, CA 95008, USA; 2 Southern California Edison Company, Corporate Environment, Health & Safety, 2244 Walnut Grove Avenue, Rosemead, CA 91770, USA; 3 Exponent, Health Sciences Group, 149 Commonwealth Dr Menlo Park, CA 94025, USA; 4 Electric Power Research Institute (EPRI), PO Box 10412, Palo Alto, CA 94303, USA
* Author to whom correspondence should be addressed. Tel: 001 650-855-1061; fax: 001 650-855-1069; e-mail: rkavet{at}epri.com
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Exposure to contact current may occur when the body is in contact with two conductive surfaces with different electrical potentials. To date, no published data that describe such exposures or electrical conditions that may predispose to such exposures exist. Our investigation into contact current exposure included (i) a small sample of workers in a garment production facility with modern well-grounded equipment performing normal work tasks and (ii) a single individual simulating garment production tasks in a sewing machine repair facility with substandard equipment grounding. In both cases, we deployed a newly developed personal monitor that records contact current events at the power frequency of 60 Hz. The personal monitoring data suggested that more frequent exposure occurs in association with, and probably because of, poorer grounding practices. This preliminary conclusion was validated with controlled laboratory measurements of potentials to reference ground on specific locations of four sewing machines with different grounding characteristics. Propensity to exposure was greater in the two machines with inferior grounding characteristics, and increased in the other two when deprived of their grounding connections. Contact currents at or below threshold-of-perception levels can produce electric fields within tissues that may plausibly produce biological effects. On this basis, such exposures have been under-investigated relative to the far greater attention accorded to occupational electric and magnetic fields.
Keywords: contact current • electrical workers • exposure assessment • exposure guidelines • sewing machines
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This paper addresses the electrical conditions under which sewing machine operators may be exposed to electrical current from touching a sewing machine, i.e. exposed to ‘contact current’. Contact current exposure occurs when a person is physically in contact with two conductive objects that are at different electrical potentials. The amount of current that flows into a person depends on the open circuit voltage of the source, as well as that person's electrical impedance relative to the source impedance. An individual's impedance is a function of the sites of contact and thus the current's path (e.g. hand to hand), the moisture content of the skin at the point of contact, the individual's physical dimensions and tissue composition (e.g. fat versus muscle). Moisture content is quite important, as dry skin has high resistivity, though it is not an absolute insulator. The variables influencing current flow have been excellently reviewed by Reilly (1998)
.
Although occupational power frequency (60 Hz in the US) magnetic field exposure has been extensively evaluated, especially in the electric utility industry, as well as electric field exposure, though to a lesser extent (Sahl et al., 1994; Savitz et al., 1994; Bracken and Patterson, 1996; Bracken et al., 1997, 2001; Kelsh et al., 2000), data concerning occupational contact current exposure are virtually nonexistent. The International Commission on Non-Ionizing Radiation Protection (ICNIRP) and the Institute for Electrical and Electronic Engineers (IEEE) have published guidelines that recommend exposure limits for electric and magnetic fields, as well as for contact currents in both public and occupational environments (ICNIRP, 1998; IEEE, 2002). The occupational contact current limits—1.0 mA for ICNIRP and 1.5 mA for IEEE—are specified to prevent aversive or painful perception. Contact currents also have recently been under investigation as a residential exposure that may underlie the association between magnetic fields and childhood leukemia (Brain et al., 2003; Kavet, 2005).
Health concerns for sewing machine operators were raised by Sobel and co-workers who focused on garment workers as a group possibly at risk of neurodegenerative disease because of magnetic field exposure from their sewing equipment (Sobel et al., 1995, 1996; Davanipour et al., 1997). More recently, women garment workers were considered a possible target population for studying breast cancer in relation to magnetic field exposure (Goodman et al., 2002; Kelsh et al., 2003). With respect to the latter investigations, sewing machine operators were initially identified as a potential high exposure group for magnetic fields and an occupational group where women made up the majority of the workforce and could be more efficiently studied for female-specific health outcomes such as breast cancer.
At a sewing machine, contact current may flow from hand to hand, if each is on the machine at points of different electrical potential or from hand-to-foot as illustrated in Fig. 1. Recently, a device for monitoring personal exposure to contact current has been developed (Niple et al., 2004). We took advantage of this new technology to conduct measurements at two commercial garment industry facilities, and to understand better the discrepancy between the results from these two sites by conducting laboratory simulations.
|
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Contact Current Meter
Briefly, the Contact Current Meter (CCM) system monitors electrical potential (i.e. voltage) differences between four medical skin electrodes that are placed on the extremities of the test subject, i.e. one placed at each of the subject's wrists and ankles. The electrodes are standard medical devices commonly used for several different types of health monitoring. The electrodes make electrically conductive connections to the skin and employ an adhesive to remain in place for extended monitoring periods. A shielded cable assembly connects the electrodes to a battery operated data acquisition meter, which is carried by the test subject and automatically records contact current events. When the potential difference across a pair of electrodes exceeds a user-preset threshold, the CCM triggers, meaning it records the voltage signal in real time, with acquisition of the signal continuing over a 512 ms period. The differences in potential are then converted into contact current (see the description below).
In order for the CCM to operate properly, the potential difference between the body of the subject wearing the meter and the voltage reference point within the meter (the CCM's local ground) cannot deviate from a tightly controlled range. To maintain this control the CCM introduces a very small current (0–1.67 µA) through a ‘bias’ cable extending from the CCM to a nearby point on the body. The potential difference on the bias channel varies with several environmental characteristics, the most significant being ambient space potentials (or electric fields). The meter may thus trigger on either contact currents flowing through a subject's body or on ambient electrical potentials monitored in the bias channel. Regardless, the filtering algorithm, summarized below, ‘decides’ if the trigger event included an exposure to a 60 Hz contact current.
Quite often static potential differences develop on adjacent objects through accumulation of charges (e.g. a cloth rubbing a plastic comb). Static discharge may occur between a person and a nearby object (a light switch) or quite commonly between a person's clothing and body below the level of perception. The CCM, although designed to detect power frequency contact currents, also triggers on static charge dissipation, referred to simply as electrostatic discharges (ESDs). Again, the filtering algorithm classifies these events into spurious discharges or signals with embedded 60 Hz contact currents (see below).
The conversion of voltage measurements into electrical currents between electrodes is computed based on impedance estimates for each contact current pathway. The current computation does not rely on external skin contact resistance, which has variability that may span four orders of magnitude (Reilly, 1998), but rather on the point to point impedance pathway beneath the superficial skin surface. The impedance value is variable from person to person, depending on physical dimensions and tissue composition. It can be estimated with default values based on height and weight or by calibrating each subject with a known current (10 µA). These techniques can reduce variability to ±20% (Niple et al., 2004).
To discriminate ESD events from power frequency contact currents, a post-processing algorithm was introduced in the CCM software to evaluate the presence of 60 Hz current following each trigger. Basically, the acquired signal is compared with a pure 60 Hz sinusoid by a correlation analysis. Based on the correlation result and the amplitude of the 60 Hz component of the acquired signal relative to the amplitude of other frequency components, the program recognizes the trace as a legitimate 60 Hz event or not. Although the detection criteria are somewhat arbitrary, they can identify 60 Hz signals that are otherwise undetectable by a visual examination. Table 1 presents the CCM's specifications. For more detail see Niple et al. (2004).
|
Measurements in commercial garment industry facilities
Personal exposure measurements of garment workers were initially performed at a garment production facility in Ventura, California in December 2001. This facility was a very modern, well-maintained work environment with mostly new equipment (computer controlled, DC machines) that appeared to be well-grounded. After employer consent and approval by facility health and safety staff, we discussed the survey purpose and provided a demonstration of the CCM to the facility employees. Potential volunteers were informed of the health and safety review that was conducted by the local electric utility company and the fact that this meter was used in a measurement survey among utility workers (Bowman et al., submitted for publication). Volunteers were then recruited for participation in the CCM survey. Employees were informed through their management and the research team that their participation was completely voluntary. Four different meters were used to record a total of
13 h of measurements among seven different workers. While wearing the CCMs, participants were asked to perform their normal work tasks. Each of the seven workers wore a CCM for up to 2.5 h of their workday. Trigger levels for the garment production facility were set at 20–75 µA peak contact current for the majority of the contact current paths monitored. The second garment industry business where personal exposure measurements were performed was at a sewing machine repair and sales facility in Los Angeles. This site was selected because of its large inventory of older sewing machines, which represented the technology used in the 1970s and 1980s in larger facilities and may still be used today in smaller, less modern facilities. The company repairs and refurbishes older sewing machines, which are then sold or rented to other businesses in the surrounding Los Angeles garment district.
The repair and show room in Los Angeles is located in a large ground floor room, which appears to be a former retail facility. Due to the lack of electrical outlets near the machines, power was supplied by a long extension cord from the back wall of the building. In November 2002, we obtained data for our operator (J. D. Sahl) at four machines under five operating conditions. The operator simulated typical sewing machine use by a garment worker. The operator wore street clothing with no special arrangements for grounding. The first two operating conditions were intended to represent background exposure while simply sitting for 1–2 min by the machines, and the last three represented machine work of 2 min duration, as follows:
- Ambient—the sewing machine under test was not connected to an electrical outlet; however, the room lights were on and an extension cord was connected to an electrical outlet at the back wall.
- Power Off—the sewing machine under test was connected to an electrical wall outlet through an extension cord, but the power to the machine was turned off at the machine.
- Idle—sitting idly at a sewing machine with the Power On but the motor not running.
- Power On/Motor Off—sitting at a sewing machine with the Power On but the sewing mechanism not running while touching the machine in various simulated work activities.
- Full Operation—sitting at a sewing machine with the Power On, the sewing motor and mechanism running while touching the machine in various simulated work activities.
Since very few power frequency contact currents were recorded at the garment production facility in Ventura (see Results), the threshold settings for the CCM were lowered to 4 µA peak contact current (and low levels of bias voltage) for the Los Angeles machine repair and sales facility. With these settings the CCM collected almost continuous data, which allowed us to examine a larger sample of possible contact currents at this facility than were observed at the Ventura facility.
Laboratory simulations
The measurements at the two facilities led us to suspect that despite different trigger levels on the CCM for the two facilities, contact current exposure would occur more often with equipment with substandard grounding characteristics either due to their older design or lack of maintenance (see Results). We, therefore, obtained four older, previously used sewing machines to test whether grounding characteristics would lead to exposure susceptibility. Two of these machines, which we designated A and B, were models similar to those tested in the repair facility showroom in November 2002 (and possibly the same machines). These machines operated with direct current (DC) motors and were computer controlled, but with different configurations. The other two sewing machines, C and D, were generally similar in design and condition to the sewing machines tested at the repair facility, but used an alternating current (AC) motor. The machine characteristics are summarized in Table 2.
|
The potential of these four machines to cause contact currents was evaluated by measuring the voltage across resistors inserted in possible pathways from the equipment to ground reference under a variety of machine operating conditions. Voltages were measured at two loads (10 M
and 1 k) to estimate exposures that could occur with a well insulated and a reasonably well-grounded person, and contact current computed using Ohm's law. The closest source of a ‘low resistance’ earth ground reference was a freestanding steel support post along one wall of the laboratory extending from floor to ceiling, and all measured test voltages were made relative to this object. The relevant sites for electrical settings (e.g. switches) and possible contact current exposure are identified in Fig. 2 for a ‘typical’ sewing machine. Using this figure as a guide, each of our four machines was examined for the presence or absence of grounding connections at each site, with the observations summarized in Table 3. As noted, only machines C and D had continuity to grounding pin connections on the ON/OFF and electric motor power cords and only machine B had a ground connection for the electric motor of the sewing mechanism. The absence of these grounding connections, which can lead to a hazardous shock, most likely reflects the disrepair of these machines associated with their age.
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Garment production facility
During the
13 recorded hours of data collection over a 2-day period at the Ventura facility, 515 triggering events occurred. Only one of these was caused by a power frequency contact current that exceeded the threshold trigger. For this event, the 60 Hz contact current reached a peak of 60 µA root-mean-square. The other 514 events were triggered by ESD. We used the post-processing algorithm described in Methods to determine the fact that embedded in 20 such traces (4%) was clear evidence of coexisting 60 Hz exposures in the range of 1–5 µA. It, thus, appeared to us that power frequency contact currents at this prototype garment production facility were rare and generally of low amplitude.
Sewing machine repair facility
During the measurement period at the sewing machine repair facility, 616 events were triggered, with 610 (99%) containing 60 Hz, using the post-processing criteria. We estimate that as many as 25% of these were triggered by contact currents exceeding the preset thresholds of 4 µA. The remaining triggers were caused by ESD events that exceeded the CCM contact current thresholds or by triggering signals on the bias channel. We considered three current pathways: Right Arm to Left Arm (RA–LA), Left Arm to Left Leg (LA–LL) and Right Arm to Right Leg (RA–RL). For the Ambient condition, 60 Hz exposures were very low for all pathways, rarely exceeding 0.5 µA. For the Power Off condition, Fig. 3 displays a family of CCM traces for all pathways for a computer-controlled machine with a DC motor and foot pedal control. Above the ambient condition this machine displayed an increase in exposure for Power On/Motor Off. For this machine, Fig. 4 shows the pathways with the highest and lowest exposure levels. The maximum level observed in all cases for this machine was 12 µA. A second machine (manual operation, AC motor with foot pedal control) displayed lower exposures, <2 µA for 90% of the triggers, with occasional exposures from 20 to 165 µA when contacting the machine with the Power On/Motor Off. While simulating work activities with the machine on (Full Operation) we observed peaks of 15–20 µA for the RA–LA and RA–RL pathways. Finally, the third machine used to observe Idle, Power On/Motor Off and Full Operation conditions (manual operation, AC motor with foot pedal control) had uniformly low exposures, with 95th percentile levels never exceeding 3 µA for any operating condition–path combination and no exposure ever exceeding 4 µA.
|
|
Laboratory simulations
Laboratory simulations were conducted to more definitively identify the electrical conditions under which exposure to contact current would be more likely to occur. To represent the range of current pathways that may exist when an operator is in contact with a machine, simulations were conducted for low resistance (1 k
) and high resistance (10 M) current pathways between a point of possible contact on the sewing machine and the reference ground in the laboratory. For machines A and B, the motor could be off with the power on; for machines C and D, the motor was automatically on when the power was on. C and D were tested ‘as is’ with their grounding connections intact (Table 3) and with these connections removed (‘ungrounded’), the latter was accomplished by simply using a 3-prong to 2-prong adapter at the wall's power socket. C and D did not have control switches. Estimated contact currents are shown in Tables 4 and 5.
|
|
In the Power Off mode, negligible exposure levels were recorded for machines A and B, but possible low resistance exposures were noted on machines C and D in association with the electric motor, the lamp, the On/Off Switch and the table frame (C only) (Tables 4 and 5, upper panel). For both ‘Power On, Motor Off’ and ‘Power On, Sewing Mechanism On’, machines A and B showed evidence of potential exposure at the control switches, electric motor, sewing mechanism and table frame (Table 4, middle and lower panel). These levels ranged from single digit microamperes for the high resistance pathway to hundreds of microamperes for the low resistance pathway, peaking at 650 µA for machines A and B at the electric motor. For machines C and D in ‘as is’ conditions with ‘Power On, Motor Off’ and ‘Power On, Sewing Mechanism On’ (Table 5, middle and lower panel), very low exposure levels were recorded for the high resistance pathway with several readings exceeding only 1 µA. For the low resistance pathway these rose, with levels at the electric motor and On/Off Switch at or around 10 µA. Thus, in general, while the better grounded machines, C and D, had lower current levels in the ‘Power On, Motor Off’ and ‘Power On, Sewing Mechanism On’ modes, these same machines unexpectedly had greater, though still low, currents than A and B in the ‘Power Off’ condition.
When machines C and D were deprived of their grounding via the power cord, currents at all sites except the lamp increased, although the low resistance current levels did not rise to the levels observed for machines A and B, with the highest (between 45 and 75 µA) recorded for the electric motor and the On/Off Switch. Table 5 shows ‘as is’ and ‘ungrounded’ comparisons for machines C and D for two pathway resistances, three operating conditions and five machine locations, for a total of 60 sets of comparisons. The advantage of C and D in producing smaller currents, especially during ‘Power On, Motor Off’ and ‘Power On, Sewing Mechanism On’ is reduced when the power source for these machines is not properly ground. For 47 of the 60 comparisons current increased in the ungrounded condition. Out of 60, 12 were unchanged, and 1 showed a decrease. Of the 12 unchanged comparisons, 2 were measurements of 0 µA where the measurement equipment was not capable of discerning a difference. The remaining 10 were for the lamp measurement site, which appeared to be unaffected by a grounded power source. The one of 60 comparisons where contact current decreased with an ungrounded power source was rechecked; however, no explanation could be found.
In several instances during the laboratory simulations (e.g. control switches for machine A), current was greater for the 10 M resistor than for the 1 k resistor, an unexpected observation that we checked, but could not be readily explained. This phenomenon, no doubt, involves other electrical pathways in the machines and possibly other non-linear impedances associated with the sewing mechanism and its control.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Equipment grounding is a critical element of electrical safety. Proper grounding ensures that in the event of a fault condition (i.e. a short circuit), equipment operators will not be exposed to a dangerous voltage through their contact with a conductive part of the equipment's exterior. Such contact could lead to serious shock and consequent electrical injury. Even in the absence of a fault, an operator may provide a pathway for current due to small potential differences between the equipment chassis and the ground (Fig. 1).
Occupational exposure to power frequency contact currents has not previously been documented, and data acquired in the field with the CCM have not been previously reported. Garment workers constitute a population in frequent manual contact with electrical equipment. We sought to assess the propensity of these workers to contact current exposure, and the attendant role of equipment grounding.
We considered that the widely different CCM contact current trigger thresholds used in the garment production facility (
20 µA) and the repair facility (4 µA) could pose a major issue with respect to the validity of our early observations and conclusions. The observation of only a single supra-threshold 60 Hz trigger in the garment facility suggested that high level contacts do not occur frequently there. In this facility, 20 ESD events were accompanied by low-level (<5 µA) 60 Hz contacts, whereas almost 500 ESD events were devoid of detectable 60 Hz currents. In the repair facility, virtually every ESD event had 60 Hz content, leading us to conclude that these events included contact current at some level. It remains possible, however, that the trigger differences led to the diverse observations in the two facilities. Our provisional conclusion that the difference in contact currents had more to do with equipment grounding (as determined by visual inspection of the facilities) than with the different CCM thresholds was our justification for testing the effect of grounding in the laboratory experiments.
The laboratory measurements, indeed, helped to elucidate the role that equipment grounding plays in affecting potentials that could lead to contact current exposure. The measurements in the repair facility, where the grounding conditions were questionable, resulted in currents ranging from <1 to >100 µA; however, the preponderance of currents were <10 µA. The measurements in the laboratory took into account that the resistance of the contact current pathway may vary over four orders of magnitude. The range of exposure levels inferred from the laboratory measurements were consistent with those logged on the CCM in the garment repair facility. The exposure levels in the repair facility appeared to be more consistent with the higher resistance pathway most likely reflecting the high impedance associated with footwear; this impedance would probably consist of parallel resistive and capacitive pathways. Whether resistive or capacitive, the total current would be logged on the CCM. When machines C and D were deprived of their grounding connections, their contact potentials noticeably rose. In contrast, the garment production facility, equipped with modern and presumably well-grounded equipment, had a low occurrence of 60 Hz contacts with all except a single record at 60 µA remaining in low single digit microampere levels. In no case at the production facility, the repair facility or in the laboratory did any exposure level exceed the ICNIRP or IEEE occupational exposure limits (1.0 and 1.5 mA, respectively). Although no health effects per se have been attributed to such exposures, we believe that they deserve attention from a dosimetric perspective. Exposures of 100 µA, though not perceptible by an adult (ICNIRP, 1998), can produce electric fields of 400 mV m–1 or more in portions of the extremities (Dawson et al., 2001). Given that biological effects may plausibly result from tissue doses as low as 1 mV m–1 (NIEHS-Working-Group, 1998), contact currents may be worthy of the level of study that has been accorded to power frequency electric and magnetic fields. An ongoing study has collected CCM and coincident magnetic field exposure data in 80 shifts of low-voltage electric utility work, with results expected in 2005 (Bowman et al., submitted for publication).
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Inadequate grounding may be a cause of exposure to contact current in workplaces beyond the garment industry. Propensity for exposure from energized equipment depends on various factors (e.g. clothing, moisture), including the manner in which the equipment is grounded. With improved grounding practices associated with modern equipment and improved facility maintenance, exposures to contact current may now be lower across segments of the garment industry. However, poor grounding, aging and poorly maintained equipment may contribute to exposures in the existing facilities.
Although voltage measurements in a laboratory provide data on equipment sites that might contribute to exposure, live monitoring with the CCM provides an improved method for capturing actual exposures. Industrial hygienists associated with industries in which workers are in manual contact with electrical equipment may wish to consider sampling potentials at possible sites of contact, with a range of resistances (and capacitance, if appropriate) to ground (e.g. 1 k–10 M). Should sites of possible exposures be identified, further investigation into the range of possible exposure levels may be advisable, and the CCM may prove a useful instrument in such pursuits.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This work was supported by EPRI projects 006778 and 006269.
Received March 3, 2005; in final form May 23, 2005
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Bowman JD, Niple JC, Kavet R. Pilot measurements of ELF contact currents with electric utility workers. Submitted.
Bracken TD, Patterson RM. (1996) Variability and consistency of electric and magnetic field occupational exposure measurements. J Expo Anal Environ Epidemiol; 6: 355–74.[Web of Science][Medline]
Bracken TD, Senior RS, Rankin RF et al. (1997) Magnetic field exposures in the electric utility industry relevant to occupational guideline levels. Appl Occup Environ Hyg; 12: 756–68.
Bracken TD, Rankin RF, Senior RS et al. (2001) Magnetic-field exposures of cable splicers in electrical network distribution vaults. Appl Occup Environ Hyg; 16: 369–79.[Medline]
Brain JD, Kavet R, McCormick DL et al. (2003) Childhood leukemia: electric and magnetic fields as possible risk factors. Environ Health Perspect; 111: 962–70.[Web of Science][Medline]
Davanipour Z, Sobel E, Bowman JD et al. (1997) Amyotrophic lateral sclerosis and occupational exposure to electromagnetic fields. Bioelectromagnetics; 18: 28–35.[CrossRef][Web of Science][Medline]
Dawson TW, Caputa K, Stuchly MA et al. (2001) Electric fields in the human body resulting from 60-Hz contact currents. IEEE Trans Biomed Eng; 48: 1020–6.[CrossRef][Web of Science][Medline]
Goodman M, Kelsh M, Ebi K et al. (2002) Evaluation of potential confounders in planning a study of occupational magnetic field exposure and female breast cancer. Epidemiology; 13: 50–8.[CrossRef][Web of Science][Medline]
ICNIRP. (1998) Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). International Commission on Non-Ionizing Radiation Protection. Health Phys; 74: 494–522.[Web of Science][Medline]
IEEE Standards Coordinating Committee 28. (2002) "IEEE Standard for Safety Levels with Respect to Human Exposure to Electromagnetic Fields, 0–3 kHz." Institute of Electrical and Electronics Engineers, Inc., IEEE Std C95.6TM-2002, New York, NY.
Kavet R. (2005) Contact current hypothesis: summary of results to date. Bioelectromagnetics (in press).
Kelsh MA, Kheifets L, Smith R. (2000) The impact of work environment, utility, and sampling design on occupational magnetic field exposure summaries. Am Ind Hyg Assoc J; 61: 174–82.
Kelsh MA, Bracken TD, Sahl JD et al. (2003) Occupational magnetic field exposures of garment workers: results of personal and survey measurements. Bioelectromagnetics; 24: 316–26.[CrossRef][Web of Science][Medline]
NIEHS-Working-Group. (1998) Assessment of health effects from exposure to power-line frequency electric and magnetic fields. Working Group Report. Research Triangle Park, NC: National Institute of Environmental Health Sciences.
Niple JC, Daigle JP, Zaffanella LE et al. (2004) A portable meter for measuring low frequency currents in the human body. Bioelectromagnetics; 25: 369–73.[CrossRef][Web of Science][Medline]
Reilly JP. (1998) Applied bioelectricity: from electrical stimulation to electropathology. New York: Springer-Verlag.
Sahl JD, Kelsh MA, Smith RW et al. (1994) Exposure to 60 Hz magnetic fields in the electric utility work environment. Bioelectromagnetics; 15: 21–32.[CrossRef][Web of Science][Medline]
Savitz DA, Ohya T, Loomis DP et al. (1994) Correlations among indices of electric and magnetic field exposure in electric utility workers. Bioelectromagnetics; 15: 1193–204.
Sobel E, Davanipour Z, Sulkava R et al. (1995) Occupations with exposure to electromagnetic fields: a possible risk factor for Alzheimer's disease. Am J Epidemiol; 142: 515–24.
Sobel E, Dunn M, Davanipour Z et al. (1996) Elevated risk of Alzheimer's disease among workers with likely electromagnetic field exposure. Neurology; 47: 1477–81.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||