Bioelectromagnetics 25:102^106 (2004)

Importance of Addressing National ElectricalCode1ViolationsThat Result in Unusual

Exposure to 60 Hz Magnetic Fields

Jack Adams,1* J. Samuel Bitler,1 and Karl Riley2

1Merrimack College, NorthAndover, Massachusetts2ELFMagnetic Surveys, St. James City, Florida

We evaluated wiring in multifamily developments containing National Electrical Code1 (NEC1)violations as a source of unusual exposure to 60 Hz magnetic fields. Two methods were used in thisevaluation: measurement and modeling. We measured the building wiring as a source of magneticfields in six multifamily developments inMichigan. In this small sample, buildingwiring proved to bean important source of exposure in four of the six cases. In all four cases with exposure from buildingwiring, one or more NEC violations were involved. To supplement our measurement efforts, we usedcomputer modeling to compare magnetic field exposure due to building wiring with magnetic fieldexposure fromexternal power lines.Our calculations showed thatwhere the buildingwiring has aNECviolation leading to net current loops, the exposure due to wiring is likely to be more important thanthat from external power lines. Our results support the results obtained in a recent study of the exposureofCalifornianK-12 students tomagnetic fields,where buildingwiringwith one ormoreNECviolationwas found to be the single most important exposure source. If 60 Hz magnetic fields are important toavoid, then improved enforcement of theNEC, as required by law, is perhaps the singlemost importantmitigation policy to adopt. Bioelectromagnetics 25:102–106, 2004. � 2004 Wiley-Liss, Inc.

Key words: ELF; exposure source; policy

INTRODUCTION

The question of whether exposure to 60 Hzmagnetic fields from sources ranging from appliancesto transmission lines has any significant health impactremains a controversial one, even though more than25 years have elapsed since the first US studies lookingat childhood leukemia were carried out [Wertheimerand Leeper, 1979]. Many laboratory studies in the ELFrange have reported no significant effects or equivocaleffects that are hard to interpret, although some in-vestigators have reported substantial in vitro [e.g.,Blackman et al., 2001; Harland and Liburdy, 1997] andin vivo [e.g., Litovitz et al., 1994] effects at intensitylevels comparable to typical environmental exposures.Epidemiological studies have also provided mixedresults, with two recent evaluations of pooled results forthe major studies to date indicating a statisticallysignificant relationship between residential proximityto transmission lines and childhood leukemia [Ahlbomet al., 2000; Greenland et al., 2000]. The backdrop ofinconsistent in vivo and epidemiological studies hasresulted in a number of suggestions as to what might beappropriate policy, including prudent avoidance [Mor-gan, 1995] and the precautionary principle [Jamiesonand Wartenberg, 2001].

The major ELF research initiative carried out inCalifornia over the past 8 years included a study inwhich the ELF exposure of children in K-12 publicschools was assessed [Zaffanella, 2000]. In this study,all school sources excepting appliances were consid-ered; included were transmission and distribution lines,electrical panels, and buildingwiring. Perhaps, themostsurprising result of this exposure study is that buildingwiring is found to bemuchmore important than all othersources evaluated. Exposure due to the wiring wasmostly found to result from National Electrical Code1

[National Fire Prot. Assoc., 1999] violations.The purpose of this study is to look further at the

question of the importance of buildingwiring as an ELF

�2004Wiley-Liss, Inc.

——————Grant sponsor: Michigan State Housing Development Authority.

*Correspondence to: Jack Adams, Merrimack College, 315Turnpike St., North Andover, MA 01845.E-mail: Jack.Adams@Merrimack.edu

Received for review 22 April 2002; Final revision received 15May 2003

DOI 10.1002/bem.10155Published online in Wiley InterScience (www.interscience.wiley.com).

source and at the impact of NEC violations. If buildingwiring is dramatically more important than externalsources, as suggested by the California school study[Zaffanella, 2000], then building wiring needs to begiven proper attention relative to other ELF sources.Weemphasize that in this study, we conducted measure-ments on multifamily dwellings, and our results andconclusions apply to multifamily and not single familydwellings.

MATERIALS AND METHODS

There are two parts to the methodology of thisstudy, measurement and computer modeling. Wemeasured the ELF magnetic fields in six Michiganmultifamily developments. The computer modelingsimulated the ELF exposure from transmission linesand in rooms affected by wiring with NEC violationsresulting in net current loops. Before describing themethodology, we provide background information onhow NEC violations lead to unusual ELF exposures.

How Violations Can Result in ELF Fields

In typical, correctly wired building circuits the‘‘hot’’ and the return currents are equal, and the 60 Hzfields from the conductors start out low and drop offrapidly. By contrast, building circuits containing whatwe term ‘‘net current’’ paths due to NEC violations arethe source of ELF fields significantly greater thanbackground.We define ‘‘net current,’’ when referring towiring, as the unbalanced resultant current carried by acircuit when some of its neutral return current has beendiverted to another circuit or conducting path. It alsoapplies to the current on the diverted path, which can bemetallic paths such as water pipes or another circuit’sneutral conductor. In a properly wired circuit, all of thecurrent is on conductors that are in close proximity, andthe circuit has no ‘‘net current.’’ Net currents are potentbecause they create a highmagnetic field adjacent to theconductors that weakens only slowly, proportional to asone over the distance (d, in meters). The magnetic fieldB (mT) from a line current source i (A) is B¼ 0.2i/d. So5 A net current produces 1.0 mTat 1 m from the circuit.

There are several NEC violations that can lead tounusual ELF fields, and each violation has the sameresult: some neutral current in a circuit returns by apath other than its dedicated circuit, and thus netcurrents are present. The most common situation seenin theCalifornia schoolmeasurement study [Zaffanella,1999]waswhere two circuits encircling a large roomonopposing sides come together in a common junctionbox, and the electrician connected the neutrals from thetwo circuits, resulting in net currents in both circuits.An indepth examination of code goes well beyond the

scope of this study; further details and references aregiven in [Riley, 1995].

Field Measurements of MultifamilyDevelopments

This portion of the work was done in conjunc-tion with the Michigan State Housing DevelopmentAuthority (MSHDA). Six developments were selectedwith an attempt to have a range of construction stylesand state of occupancy represented. Five sites weremultilevel, usually four or so levels, and one site wastownhouse style with separate units. Four structureswere not yet occupied, but had energized services sothat lights and appliances could be turned on. Two siteswere near or within Detroit, and the others were within100 miles of Detroit. The measurements all took placeduring July, 2000.

The protocol for a site visitwas first to determine ifthere were significant fields from nearby power linesthat would affect the fields within the building. Allmeasurements were made with a Bell 4090 3 axisGaussmeter. In no case did power line fields have asignificant impact. Awalk through of the building wasthen conducted to determine if unusual (>0.05mTor so)ELF fields were present. In cases where unusual fieldswere present, the circuit or circuits that had net currentscausing the fields were traced down. An attempt wasthen made to determine precisely what NEC violationcaused the net currents by tracing down the source of thenet currents in each case.

Computer Modeling

During the course of this work we developed twocomputer models, one to evaluate the ELF field profilenear a transmission line and another to evaluate the ELFfields in a room that has net currents in one or morewalls. The software package used to implement themodels is MATHCAD, and the transmission and theroom models are based on the Biot–Savart law. Toevaluate the field profile near a transmission line, theinputs to the MATHCAD document are the location,loading, and phase of all conductors. A simplifiedmodel is utilized, where sag is neglected and the linesare assumed straight. Using a similar MATHCADdocument, fields in a rectangular roomwith net currentsin up to four walls are calculated. The required inputsare the room length and width, and the location, magni-tude, and phase of all net currents.

In this study we looked at field profiles near twohypothetical, single circuit transmission lines, one a69 kV line with 300 A per phase and the second a230 kV line with 600 A per phase. A flat, vertical wireconfiguration was assumed, with 1.83 m betweenconductors for the 69 kV line and 3.66 m for the

National Electrical Code1Violations 103

230 kV line. Configurations and loading can varydramatically. In our case they were based on industrydata and chosen so as to yield typical field valuesadjacent to the lines.

A straightforward hypothetical net room currentcase was evaluated, where a net current of 5.0 A wasassumed in all four walls. The figure of 5.0 Awas takenas being a reasonable value, based on net currentsobserved during the measurement portion of this studyand on typical values encountered by one of the authors(KR) doing hundreds of net current measurementsacross a broad range of structure types between 1990and2002.Theconductorswere assumed tobe 2mabovethe floor, and exposure was calculated at 1 m above thefloor.

RESULTS

Field Measurements ofMultifamily Developments

The primary finding in the measurement part ofthis study is that in four of the six housing develop-ments, significantly elevated EMFs due to NEC viola-tions were measured. In two cases, the industrial sizeddryers for building occupants had been incorrectlyinstalled. The manufacturer ships the units with theneutral and equipment wires connected internally toallow for installations in buildings with older 3 prongoutlets. However, when installing in a new building,where 4 prong outlets are required by code for industrialdryers, the contractor must remove the internal neutral/grounding connection according to the owner’smanual.When the contractor fails to disconnect this internallink, the shorting of neutral and ground creates an alter-nate path for neutral current, even when the machinesare not turned on. We conducted measurements in thehallway near the laundry room and watched the ELFfields incrementally decrease as we sequentially un-plugged the machines, which were not in use. Con-versely, turning the machines on caused the fields torise.

In another multifamily development, 3 wayswitches were wired with 2 wire Romex inside themain utility room, a clear code violation resulting insignificant net currents. Fields exceeded 0.4 mT in theunit above the utility room. The building’s owner sub-sequently had those switches rewired, virtually elim-inating the fields due to that cause. In a fourth case, netcurrent at most subpanels indicated a code violation,although the precise location was not determined andcould possibly have been behind drywall. In this fourthcase, the conductors left the subpanels in a bunch ratherthan individually, which was a code violation for a flush

mounted subpanel and which made diagnosis moredifficult.

In two housing developments, NEC violationsleading to net currents were not seen. One case was atownhouse style development, within which each unithad its own electric service. Another case was a multi-floor development that had no net currents encounteredduring our visit.

To summarize the results of this portion of ourstudy, in four of six sites unusual ELF fields due to netcurrents were measured. In the four cases NEC viola-tions were located, and in one of those cases there wasevidence of still further violations, which we were notable to find in the time available. Of the two structuresthat did not have unusual fields, one was the onlytownhouse development surveyed. Townhouse devel-opments might be expected to avoid some of the netcurrents seen in the other developments, as there are nolarge utility rooms, and no large laundry rooms withmultiple industrial grade dryers.

Computer Modeling

The calculated 60 Hz magnetic field profiles fortwo hypothetical, single circuit transmission lines, oneat 69 kVand the other at 230 kV, are shown in Figure 1.Note that for the 69 kV line, the calculated fields havedropped to 0.4 mTat 20 m from the line and to 0.2 mTat30 m from the line. For the 230 kV line, the calculated

Fig. 1. Field strength in mT vs. distance from transmission line.Upper curve is 230 kV linewith 600 A/phase, lower curve is 69 kVlinewith 300A/phaseassumed.

104 Adams et al.

fields have dropped to 0.4 mTat 50 m from the line, and0.2 m at 70 m from the line.

The calculated fields for a room with 5.0 A netcurrents in all four walls are shown as a contour plot inFigure 2.Note that the calculated fields are above 0.3mTthroughout the roomand exceed 0.75mTin about half ofthe room.

DISCUSSION AND CONCLUSIONS

The measurement portion of our study, while itis it is a very small sample, is along the lines of theCalifornia study result [Zaffarella, 2000] that net cur-rents are an important exposure source in 20% of class-rooms. In addition, we did not see any examples of ELFfields from power lines impacting exposure inside thestructures visited, which also is along the lines of theresults seen in the California study.

The ELF field calculation part of the study givesinsight into why net currents show up as an exposuresource more often than power lines. Fields directlyunder power lines can be fairly high, but drop offroughly as 1/R2, while in a room enclosed by netcurrents the fields drop off as 1/R or even more slowly,resulting in higher exposure where people are. When‘‘typical’’ power lines are compared to ‘‘moderate’’ netcurrents, the net currents appear to be significantlymoreimportant to address from an ELF exposure stand-

point. What we did that was not part of the CA schoolmeasurement study was to model the variation of fieldsthroughout the room.

Because in the present study, multifamily dwell-ings in Michigan have been measured, while theCalifornia study dealt with schools, our work indicatesthat NEC violations leading to net currents may becommon across many building types throughout theUnited States. We note that our results do not neces-sarily apply to single family residences, where netcurrents in nearby distribution lines and groundingcurrents have been found to be important [Zaffanella,1993].

POLICY IMPLICATIONS

We believe that there are two policy implicationsstemming from the results of this work and theCalifornia School Measurement Study [Zaffanella,2000]. First, if a comprehensive policy addressing pos-sible health effects of ELF magnetic fields is beingconsidered or adopted, fields due to building wiringneed to be included as an exposure source. Second, thepresence of ELF fields, at least in larger buildings, canoften indicate that NEC violations need to be addressedfor reasons of shock and fire hazard. This is actually nota coincidence, since the code seeks to avoid the muchhigher than necessary inductance associated with netcurrents. This increased inductance is accompanied byhigh ELFfields. It also can lead to inductive heating andcan compromise circuit breaker response.

If a policymaker chooses to address the potentialhealth effects of long term exposure to ELF magneticfields, then it appears that net currents in buildingwiringare at least as important a source to address as externalpower lines. First, it appears likely that exposure due tobuilding wiring is much more common than exposuredue to power lines. Second, a policy to deal with newconstruction is quite inexpensive and straightforward,requiring a simple additional test for electrical inspec-tors and education of electrical contractors; in caseswhere errors are discovered, the fixes are generallyquite inexpensive. By contrast, policies to deal withELF exposure from existing or even new power linesusing engineering approaches are not straightforward,involving numerous factors such as visual impact, pro-perty values, and effects on the reliability of a system.

In contrast to almost all other ELF sources, apolicymaker who places no weight at all on potentialhealth effects due to ELF fields may still choose toaddress NEC violations leading to net currents. Fromthe perspective of restricting concern to adherence totheNEC, required by both local and federal law inmanycases, unusual building wiring ELF fields typicallyindicate NEC violations resulting in net currents. The

Fig. 2. Contour plot of room15 m2, with net current of 5.0 A in allfourwalls. [Color figure canbeviewedin theonline issue, which isavailableat www.interscience.wiley.com]

National Electrical Code1Violations 105

net cost of a policy to address this particular type oferror may well be less than zero: the upfront costs arequite lowasmentioned above; and in exchange for thesemodest costs, which would become less over time aselectrical contractors learn to avoid some simple NECviolations, potential fire and shock hazards can beavoided.

A straightforward and in our assessment prudentpolicy is to include a check for NEC violations that leadto net currents as part of electrical inspections. Thepolicy advantage of avoiding NEC violations is thatshock and fire hazards are mitigated, and potentialhealth advantages are a bonus, rather than controversial.As part of any comprehensive policy seeking to dealwith the potential hazards of EMFs, buildingwiring andNEC violations need to be included.

To summarize this study, our work supports thehypothesis that buildingwiring, excluding single familyresidences, is at least as important as and probablysignificantly more important an ELF field exposuresource as are power lines. Both our limited measure-ment data and our computer modeling data support thishypothesis. From a policy perspective, NEC violationsin buildings represent a much more straightforwardsituation to address than the presence of transmissionand distribution lines, where a wide range of complicat-ing factors come into play. Finally, the addressing ofNEC violations represents one of the few cases whereadopting engineering based ELF mitigation strategyand requiring adherence to local and federal law aredirectly in alignment.

ACKNOWLEDGMENTS

The authors thank Indira Nair of Carnegie MellonUniversity, Robert Goldberg of Information Venturesfor valuable discussions. We thank the reviewers forvaluable comments. We thank Bruce Jeffries of theMichigan State Housing Authority for technical sup-

port. We thank the Michigan State Housing Authorityfor funding to carry out this work.

REFERENCES

Ahlbom A, Day N, Feychting M, Roman E, Skinner J, DockertyJ, Linet M, McBride M, Michaelis J, Olsen JH, Tynes T,Verkasalo PK. 2000.A pooled analysis ofmagnetic fields andchildhood leukaemia. Br J Cancer 83(5):692–698.

BlackmanCF,BenaneSG,HouseDE. 2001.The influence of 1.2mT,60 Hz magnetic fields on melatonin- and tamoxifen-inducedinhibition of MCF-7 cell growth. Bioelectromagnetics 22:122–128.

Greenland S, Sheppard AR, Kaune WT, Poole C, Kelsh MA. 2000.A pooled analysis of magnetic fields, wire codes, and child-hood leukemia. Epidemiology 11(6):624–634.

Harland JD, Liburdy RP. 1997. Environmental magnetic fieldsinhibit the antiproliferation action of tamoxifen and melato-nin in a human breast cancer cell line. Bioelectromagnetics18:555–562.

Jamieson D, Wartenberg D. 2001. The precautionary principle andelectric andmagnetic fields.Am JPublicHealth 91(9):1355–1358.

Litovitz TA, Montrose CJ, Doinov P, Brown KM, Barber M. 1994.Superimposing spatially coherent electromagnetic noiseinhibits fields-induced abnormalities in developing chickembryos. Bioelectromagnetics 15:105–113.

MorganMG. 1995. Fields fromelectric power:What are they?Whatdo we know about possible health risks? What can be done?Pittsburgh, PA: Carnegie Mellon University Dept. of Eng.and Pub. Policy.

National Fire Protection Association. 1999. National ElectricalCode1, First Edition. Clifton Park, NY: Delmar Learning.

Riley K. 1995. Tracing EMFs in Building Wiring and Grounding.Tucson, AZ: Magnetic Sciences International.

Wertheimer N, Leeper E. 1979. Electrical wiring configurations andchildhood cancer. Am J Epidemiol 109(3):273–284.

Zaffanella L. 1993. Survey of Residential Magnetic FieldSources, Volume 1: Goals, Results and Conclusions. PaloAlto, CA: Electric Power Research Institute. Report TR-102759-V1.

Zaffanella L. 2000. Electric and Magnetic Field Exposure Assess-ment of Powerline andNon-Powerline Sources for CaliforniaPublic School Environments, Final Report. Vols. I and II.Berkeley, CA: California Department of Health ServicesPublications.

106 Adams et al.