Full text of DOI:10.1093/oxfordjournals.rpd.a006513

Radiation Protection Dosimetry
Vol. 94, No. 4, pp. 375–380 (2001)
Nuclear Technology Publishing
TECHNICAL NOTE
OUTDOOR BACKGROUND ELF MAGNETIC FIELDS
IN AN URBAN ENVIRONMENT
G. d’Amore, L. Anglesio, M. Tasso, A. Benedetto and S. Roletti
Regional Environmental Protection Agency, ARPA Piemonte, Department of Ivrea
Via Jervis 30, 10015 Ivrea (TO), Italy
Received February 22 2000, in final revised form January 19 2001, accepted January 24 2001
Abstract — Classification of ‘exposed/non-exposed’ subjects in epidemiological studies concerning the possible cancer risk asso-
ciated with ELF magnetic field exposure is based on the a priori assumption of magnetic field value cut-off points that, often,
are defined equal to minimum exposure levels typical of a population residing near high voltage facilities (0.1–0.2 ␮T), but in
some cases an environmental magnetic field level not associated with transmission lines can exist. The results of an ELF magnetic
field survey in an Italian urban area (about 1 million inhabitants) are presented: average field levels are correlated with population
density of different districts. Exposure indexes are obtained, which are compared with those evaluated in studies regarding
domestic exposure: background average levels result in comparable to cut-off points in epidemiological studies, but in some
districts with high population density, they are much higher. This shows that knowledge of background magnetic field level in
urban areas can assume a significant role in exposure assessment.
INTRODUCTION
Karner and Stamm(4) and Preece et al(7) measured
higher residential values up to 0.22 ␮T and 0.67 ␮T
respectively.
Because of a reported excess of leukaemia among
children living near power lines and exposed to an ELF
magnetic field greater than an established cut-off level,
there is some concern that magnetic fields may affect
human health. This possibility causes considerable con-
troversy in the scientific community and receives great
attention in the exposed population.
Even if scientific foundation for limit values on low
levels of magnetic field are insufficient, a precautionary
strategy based on reducing exposure so as to reduce the
risk of injury to human beings can be adopted.
According to this precautionary strategy it is very
important to have a knowledge of the normal mag-
netic field level, that is the levels in areas where
The knowledge of outdoor background field strength
in urban areas can be of interest because there is a lack
of data about these levels where a great number of
subjects are exposed for not negligible periods.
In this work results are presented of an ELF magnetic
field survey in Turin, a large town located in the north-
west of Italy, with approximately 1 million inhabitants
occupying an area of about 130 km2. Data obtained in
the survey were used to calculate different exposure
indices commonly adopted in exposure assessment(8)
Particular attention was paid to statistical analysis.
.
human beings can be expected to be repeatedly INSTRUMENTS AND SURVEY CRITERIA
present, as pointed out in the Swedish guide ‘Low
ELF magnetic field monitoring was done by using a
commercial Emdex II survey meter. The meter fre-
quency response is in the range 40 Hz–800 Hz and its
dynamic range is 0.01 ␮T–300 ␮T. By Emdex II mem-
ory capacity (512 kb) a number of data up to 170.000
in 71 h, collecting every 1.5 s, can be logged.
Measurements were made walking with a regular
speed along an established path. The town of Turin is
subdivided into ten districts (see Figure 1) whose
population density ranges from 62 m2.inhabitants−1 to
255 m2.inhabitants−1. The ELF magnetic field was
mapped in each district along a regular grid, see
Figure 2, which is chosen to be representative of the
investigated area, in such a way that a constant ratio
between district area (A) and path length (d) was kept:
A/d Х 400 m. For example, in an area of 10 km2, a path
25 km long was walked. Data were collected each 1.5 s,
which corresponds, considering a regular walking speed
of about 4.2 km.h−1, to an interval of about 1.8 m. So
frequency electrical and magnetic fields: the pre-
cautionary principle for national authorities — Guid-
ance for Decision Makers’ (The Swedish National Elec-
trical Safety Board)(1)
.
In the great majority of epidemiological studies(2,3)
the cut-off exposure levels were defined equal to
0.1 ␮T, 0.2 ␮T or 0.3 ␮T, which are considered mini-
mum exposure levels typical of groups of population
residing close to high voltage facilities.
In reality, environmental magnetic field levels, not
due to high voltage transmission lines and comparable
with the above reported cut-off points, can exist. Several
authors(4–7) measured residential ELF magnetic fields
away from power lines; some of these found mean
values lower than a cut-point of 0.1 ␮T(5,6), while
Contact author E-mail: arpaivrea2Ȱeponet.it
375

G. D’AMORE, L. ANGLESIO, M. TASSO, A. BENEDETTO and S. ROLETTI
the number of field samples detected in each district was connected to a wheel provided with a switch that, at
12,000 on average.
each strobe, transmitted to the meter a signal for data
In a particular pedestrian area, the magnetic field as acquisition. Field spatial distribution was plotted using
a function of distance was measured by Emdex II Emcalc software.
Figure 1. Division in districts of the town in Torino.
Figure 2. Mapping of a district. The dark line is the measurement path.
376

BACKGROUND ELF MAGNETIC FIELDS IN AN URBAN ENVIRONMENT
2
The temporal variation of the magnetic field was
␴
B = exp[(ln(B−␶)] +
+ ␶
(1)
(2)
evaluated by mapping a localised area (central town) in
three periods of the day: 7 a.m.–9 a.m., noon–2 p.m.,
5 p.m.–7 p.m. which are the time intervals when most
citizen movements occur. Furthermore, a long-term
measurement in a fixed position was acquired, from
7 a.m. to 7 p.m., to estimate the daytime field variation
only as a function of changes of background source
emissions (electric energy distribution network) without
considering spatial variations.
2
BMEDIAN = exp(ln(B−␶)) + ␶
2
were ␴ is the variance of the variable ln(B-␶) and ␶ is
the detection limit value equal to 0.005 ␮T.
For all collected data, including measurements made
in all districts (size N equal to 110,667), it results that
B = 0.19 ␮T and BMEDIAN = 0.079 ␮T while the ex-
pected values calculated by the Equations 1 and 2 are
0.22 ␮T and 0.08 ␮T, respectively. This comparison
shows that Equations 1 and 2 are satisfied and, conse-
quently, that a log-normal distribution fits the survey
data.
In order to have a more accurate estimation of
statistical parameters, to include non-zero values
which cannot be measured because they are below the
detection limit, data were dealt with according the
procedure for log-normal distributions proposed by
DATA ANALYSIS AND RESULTS
For a statistical analysis of magnetic field measure-
ments, data collected in all districts were reported in a
frequency relative plot (see Figure 3). Figure 3 shows
that data are highly skewed and do not follow a normal
distribution (arithmetic mean, equal to 0.19 ␮T, greater
than the median, equal to 0.08 ␮T and greater than the
geometric mean, equal to 0.06 ␮T). So data analysis was
pointed to assess if data would fit a log-normal distri-
bution, as observed by other authors for analogous
surveys(6,9). Figure 4, where relative frequency of meas-
ured magnetic field levels versus the logarithm of field
levels is reported, shows a trend typical of a log-normal
distribution shifted to the left by an amount ␶ (three-
parameter log-normal distribution). The ␶ parameter can
be estimated as the minimum detectable value of the
instrument used for data collection.
0.3
0.25
0.2
0.15
0.1
To evaluate if a log-normal distribution fits these data
the expected relation between statistical parameters was
considered: arithmetic mean of field levels (B),
arithmetic mean of logarithm of field levels (ln B),
which can be corrected taking into account detection
limit value ␶ (ln(B − ␶)), and median of field levels
(BMEDIAN). Specifically, if the data follow a three-para-
meter log-normal distribution the following relations
0.05
0
-5.75 -4.85 -3.95 -3.05 -2.15 -1.25 0.35 0.55 1.45
lnB
Figure 4. Relative frequency of measured magnetic field levels
versus the logarithm of field levels.
should be satisfied(10)
:
20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
1
1
0.1
1.1
1.21
0.31
0.51
0.71
0.91
1.31
1.51
1.71
1.91
1.81 2.01
0.21
0.41
0.61
0.81
1.01
1.41
1.61
0.009
B (mT)
Figure 3. Frequency distribution of ELF magnetic field strength in all districts.
377

G. D’AMORE, L. ANGLESIO, M. TASSO, A. BENEDETTO and S. ROLETTI
Hornung and Reed(11). This procedure, which is valid data relevant to low density districts (Ͼ100 m2.
when the number of non-detectable values is lower inhabitants−1), data relevant to high density districts
(Ͻ100 m2.inh−1.) and those relevant to the whole
town. By the chosen criterion five districts resulted as
low density and five as high density districts. Taking
into account the district areas, 54,444 measurements
for low density districts and 56,223 measurements for
high density districts were collected. Results of the
analysis are reported in Table 1, which shows an
increasing trend of all exposure indices with the popu-
lation density, with intermediate values for the whole
town. In particular the arithmetic mean, which ranges
from 0.15 ␮T to 0.20 ␮T, is always comparable with
cut-off point values, while the median and geometric
mean are lower, according to log-normally distributed
data, and range from 0.06 ␮T to 0.08 ␮T.
Since for monitoring one district a mean period of
four hours was necessary, the field levels analysis
took both spatial and temporal field variations into
account. In order to know field variation with time, a
long-term measurement in a fixed outdoor position
was made away from substations. Statistical analysis
of the results gave a mean value of 0.05 ␮T with a
standard deviation (␴) of 0.02 ␮T and a median of
0.04 ␮T. With respect to statistical analysis made for
data collected in district monitoring, it is noted that
the dispersion of values is very low and that the
median and mean values are almost equal. These dif-
ferences indicate that data set relevant to a district
survey and long-term fixed point measurement follow
different distributions as expected because the long-
term measurement values depend only on one para-
meter (load current of distribution lines), while in the
district survey a great number of variables are
involved.
than 50% of the whole set of data and when the mode
is higher than the threshold value, conditions satisfied
by collected data, consists in replacing the limit of
detection ␶ by ␶/√2.
For each district area, these corrected data were ana-
lysed to determine the following exposure indices: arith-
metic mean, standard deviation, median, geometric
mean, maximum value and percentage of time spent
above given thresholds (0.1 ␮T, 0.2 ␮T and 0.3 ␮T).
As a consequence of experimental evidence about an
increase of the exposure levels with the population
density, data were divided into three groups for analysis:
Table 1. Exposure indices evaluated for different districts
grouped as a function of population density.
Exposure index
Low density
High density
The
(Ͼ100 m2.inh−1) (Ͻ100 m2.inh−1) whole
town area
Mean (␮T)
Standard deviat. (␮T)
Median (␮T)
Geometric mean (␮T)
Maximum (␮T)
0.15
0.24
0.06
0.04
3.3
0.20
0.30
0.08
0.11
5.73
0.18
0.27
0.07
0.06
5.73
Time Ͼ0.1 ␮T(%)
Time Ͼ0.2 ␮T(%)
Time Ͼ0.3 ␮T(%)
21.7
14.0
9.7
55.5
36.6
26.0
44.1
28.9
20.3
Table 2. Exposure indices evaluated in a central part of
town in three periods of daytime.
Variation of the field pattern with time was investi-
gated repeating the same path in a central area cover-
ing five blocks in three different day periods. The
average results are presented in Table 2, which shows
that statistical parameters increase in the central part
of the day, when there is a higher distribution of
Exposure index
7–9 a.m.
noon–2 p.m. 5–7 p.m.
Mean (␮T)
Standard deviat. (␮T)
Median (␮T)
0.15
0.24
0.06
3.3
0.20
0.30
0.08
5.73
0.18
0.27
0.07
5.73
Maximum (␮T)
cables loading(12)
Table 3. Comparison with residential magnetic field surveys.
U.K. 1994(5) Switzerland 1995(6) U.K. 1996(7) Norway 1997(13)
.
Exposure index
Italy 1997
(␮T)
H.V.
Lines
No
Lines
H.V.
Lines
No
Lines
No
Lines
H.V.
Lines
No
Lines
(survey presented
in this work)
Arithmetic mean
Geometric mean
Median
Maximum
percentage of
time Ͼ0.1
0.69
0.2
0.014–0.09
0.054–0.13
0.043–0.11
0.15
0.04
0.037
0.021
0.044
0.67
0.45
0.32
0.09
0.05
0.19
0.06
0.08
5.73
43.5
3.9
25
–
7
378

BACKGROUND ELF MAGNETIC FIELDS IN AN URBAN ENVIRONMENT
Comparison with other studies
magnetic field limit of 100 ␮T to preserve the general
public from short-term, immediate health effects.
In particular urban areas (town centre) exposure
indices reach values similar to those typical of resi-
Exposure indices adopted in this work were used
also by other authors in surveys of personal exposure
to ELF magnetic field in residences(5–7,13). A compari-
son with these studies is interesting in order to deter-
mine the contribution of the background outdoor field
to personal exposure over 24 h, but it is only indicative
because some aspects of monitoring methods adopted
by several authors were very different (sample dimen-
sion, choice of measurement points, spot measurement
or personal dosimetry, etc.). The comparison is provided
in Table 3, where we reported data relevant to resi-
dences within 100 m of high voltage lines (H.V. Line
columns) and beyond 100 m from lines (No line
columns).
dences placed within 100 m from power lines(16)
.
For a more complete evaluation of human exposure
it is significant to take into account time exposure.
Following other authors who consider time in the
exposure description(8,18,19), we can define a cumulative
dose exposure index as follows:
T
Bc =
B(t) dt
(5)
͵
0
where B (t) is the mean of the magnetic field strength
in dt time interval and T is the exposure time.
The daily average time spent outdoors by citizens,
supplied by the statistical office of Turin munici-
pality, is equal to about 80 min. If the cumulative
dose is adopted as an exposure index, assuming the
effect proportional to exposure time (even if this
hypothesis is not confirmed by epidemiological
studies(20)), the result is that the contribution of back-
ground magnetic field level to daily dose can be very
The maximum exposure levels, from 3.01 ␮T to
5.73 ␮T, and time percentage above thresholds are com-
parable to those found in ELF magnetic field surveys in
residences close to power lines(2,5,6)
.
DISCUSSION AND CONCLUSIONS
The arithmetic mean of ELF magnetic field strengths low because the outdoor time exposure (80 min) is
found in the whole town of Turin (0.19 ␮T) proved to 5% of the day. This contribution can be significant
be comparable to the cut-off point (0.2 ␮T) used in only for particular groups of workers who spend great
many epidemiological studies and adopted as a refer- part of their job time outdoor such as postmen, taxi
ence level in some regional or national regulatory drivers, road sweepers and so on.
proposals(14,15), in spite of the International Radiation
Results of this survey indicate that the knowledge of
Protection Association (IRPA) position. In fact, in urban background magnetic field levels is very
relation to a basis for limiting exposure, the IRPA important for a complete definition of general public
International Commission on Non Ionizing Radiation exposure. Mean magnetic field levels in different dis-
Protection (ICNIRP) states that ‘[. . .] results from tricts of Turin ranged from 0.15 ␮T to 0.35 ␮T, increas-
epidemiological research on EMF field exposure and ing according the density of population. These data
cancer, including childhood leukaemia, are not strong suggest that in epidemiological studies, which consider
enough in the absence of support from experimental as significant an exposure above exposure levels of
research to form a scientific basis for setting exposure 0.1 ␮T or 0.2 ␮T, background urban exposure has also
guidelines’(16). For this reason ICNIRP gives an ELF to be taken into account.
REFERENCES
1. The Swedish National Board of Occupational Safety and Health; National Board of Housing, Building and Planning; National
Electrical Safety Board; National Board of Health and Welfare; Radiation Protection Institute. Low-frequency Electrical and
Magnetic Fields: the Precautionary Principle for National Authorities – Guidance for Decision-makers (available from Statens
strålskyddsinstitut — Swedish Radiation Protection Institute).
2. Washburn, E. P., Orza, M. J., Berlin, J. A., Nicholson, W. J., Todd, A. C., Frumkin, H. and Chalmers, T. C. Residential
Proximity to Electricity Transmission and Distribution Equipment and Risk of Childhood Leukemia, Childhood Lymphoma,
and Childhood Nervous System Tumors: Systematic Review, Evaluation, and Meta-analysis. Cancer Causes and Control, 5,
299–309 (1994).
3. National Institute of Environmental Health Sciences, NIEHS. Assessment of Health Effects from Exposure to Power Line
Frequency Electric and Magnetic Fields. Working Group Report, Eds C. J. Porter and M. S. Wolfe (NC-USA) (1998).
4. Karner, H. C. and Stamm, A. R. Residential Long-term Magnetic Field Exposure. In: Proc. 7th Int. Symp. on High Voltage
Engineering, Dresda (1991).
5. Merchant, C. J., Renew, D. C. and Swanson J. Exposures to Power-frequency Magnetic Fields in the Home. J. Radiol. Prot.
14, 77–87 (1994).
6. Joss, S., Baumann, J., Kreuter, U. M., Stratmann, M. and Wernli, C. Exposure of the Swiss Population to 50 Hz Magnetic
Fields. In: Proc. 11th Int. Symp. on Electromagnetic Compatibility, Zurich, pp. 268–270 (1995).
7. Preece, A. W., Grainger, P., Golding, J. and Kaune, W. Domestic Magnetic Field Exposures in Avon. Phys. Med. Biol. 41,
71–81 (1996).
379

G. D’AMORE, L. ANGLESIO, M. TASSO, A. BENEDETTO and S. ROLETTI
8. Juutilainen, J., Hatfield, T. and La¨a¨ra¨, E. Evaluating Alternative Exposure Indices in Epidemiologic Studies on Extremely
Low-frequency Magnetic Fields. Biolectromagnetics 17, 138–143 (1996).
9. Silva, M., Hummon, N., Rutter, D. and Hooper, C. Power Frequency Magnetic Fields in the Home, IEEE Proc. 1–13 (1988).
10. Gilbert, R. O. Statistical Methods for Environmental Pollution Monitoring New York: Van Nostrand Reinhold (1987).
11. Hornung, R. W. and Reed, L. D. Estimation of Average Concentration in the Presence of Nondetectable Values. Appl.
Occup. Environ. Hyg. 5(1), 46–51 (1990).
12. Tofani, S., d’Amore, G., Fiandino, G. and Tasso, M. Assessment of Magnetic Fields Exposure from AC Power Lines in an
Italian Residential Area. In: Proc. Radiation Protection in Neighbouring Countries in Central Europe, 1995, Ljubljana.
pp. 361–364 Ed. D. Glavic-Cindro (1996).
13. Vistnes, A. I., Ramberg, G. B., Bjørnevik L. R., Tynes, T. and Haldorsen, T. Exposure of Children to Residential Magnetic
Fields in Norway: is Proximity to Power Lines an Adequate Predictor of Exposure?, Biolectromagnetics 18, 47–57 (1997).
14. Regione Veneto. Prevenzione dei danni derivanti dai campi elettromagnetici generati da elettrodotti. Legge Regionale no 27
(30 June 1993).
15. Bozza dello schema del decreto relativo ai limiti di esposizione, ai valori di attenzione e agli obiettivi di qualita` per la
tutela della salute della popolazione nei confronti dei campi elettromagnetici generati a frequenze o da impianti fissi non
contemplate dal DM 381/98 (Italian government law proposal to be approved by parliament).
16. International Commission on Non-Ionizing Radiation Protection. Guidelines for Limiting Exposure to Time-varying Electric,
Magnetic and Electromagnetic Fields (up to 300 GHz). Health Phys. 74, 494–522 (1998).
17. Tofani, S., d’Amore, G., Bonazzola, G. C. and Fiandino, G. Computational Method for Power Line Magnetic Field Evaluation
ACES J. 7, 148–161 (1992).
18. Preece, A. W., Kaune, W. T., Grainger P. and Golding, J. Assessment of Human Exposure to Magnetic Fields Produced by
Domestic Appliances. Radiat. Prot. Dosim. 83, 21–27 (1999).
19. Valberg, P. A. Designing EMF Experiment: what is required to characterize “Exposure”?. Biolectromagnetics 16, 396–
401 (1995).
20. Olsen, J. H., Nielsen, A. and Schulgen, G. Residence near High Voltage Facilities and Risk of Cancer in Children. Br. Med.
J. 307, 891–895 (1993).
380