french children exposure to 50 hz magnetic field · 1 case of emdex put close to electric appliance...
TRANSCRIPT
Isabelle Magne1, Martine Souques2, Jacques Lambrozo2, Mfoihaya Bedja3, Gilles Fleury3, Laurent Le Brusquet3, Alexandre Carlsberg4, François Deschamps5, Gilbert Belardi6
1 EDF R &D, Moret sur Loing, France2EDF Service des Etudes Médicales, Levallois-Perret, France
3 Supelec, Gif-sur-Yvette, France4 MV2, Montrouge, France
5 RTE, Paris La Défense, France6 ERDF, Paris La Défense, France
French children exposure to 50 Hz magnetic field
2 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Contents
IntroductionRecruitment of volunteers and data collectionDatabase analysisDescription and comparison of mean exposuresCharacterisation of mean exposuresElectric networksConclusion and perspectives
3 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Introduction and aim
1979: Wertheimer study
2001: classification II B of ELF magnetic field by IARC (possibly carcinogenic to human )
2007: collective assessment by international expert groups WHOStatistic association observed between childhood leukemia and magnetic field exposure higher than 0.4µT in mean over 24hNo causal relationship demonstrated
Aim of the study: what is the exposure of the French population?2007: EXPERS study initiated by the Health Ministry
4 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Recruitment of volunteers
Recruit 1000 adults and 1000 children representatives of the French population
MV2 Conseil in charge of data collectionMethod of random lotteryStart file 95 362 phone numbers (no professional)Recruitment by phone, then pollster on siteCriteria of distribution according to the distribution of the French population by region3 campaigns (February-April 2007, October 2007-April 2008 and October 2008-January 2009).
5 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Collection of data
Measurements during 24h with an EMDEX II worn by the volunteer
1 measurement every 3 sMeasurement of broadband (40-800 Hz) and harmonics (100-800 Hz)Measurement range: 0.01 to 300µTNot disturbed by GSM
Timetable filled in by the volunteerActivitiesLocationsHours
Questionnaire filled in at the end with the pollsterInformation on the volunteer (age, profession, etc..)Information on the home (year built, heating, etc...)
Measurement of GPS coordinates at the home front doorSearch afterwards of proximity of electric networks
6 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Electric networks close to home
Definition of distance “close to home”
20MV/LV substation20Underground cable LV and 20 kV20Underground cable 63 to 150 kV20Underground cable 225 kV
200Train network20Overhead line LV and 20 kV70Overhead line 63 and 90 kV
100Overhead line 150 kV120Overhead line 225 kV200Overhead line 400 kV
Distance (m)Type of network
distance
7 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Description of the database (1/2)
Summary of the phone numbers called 95 362 numbers called :
47% answered3047 agreements on principle (3%)2148 measurements performed (2,25%)
Mean time to recruit 1 volunteer = 70 min
Pertinence of data2148 measurements performed2048 measurements validated par MV2 Conseil1525 addresses (523 paired adult/child)
8 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Description of the database (2/2)
Distribution by age
Distribution by gender
49/5151/49children
databaseFrench population% M/F
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14
age in years
perc
enta
ge o
f chi
ldre
n
French populationEXPERS sample
9 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Analysis of the database
2048 measurements validated par MV2 ConseilKeying in of timetable and questionnaires
19 series deleted for different reasons2029 measurements analysed
977 measurements recorded by children (0-14 years).1 052 measurements recorded by adults (15 years and over)
Example of refined temporal cutting
1- home
2 – work on PC
3 – train
4 – meeting room
5 – bus
6 – home
7 – housework
8 - shopping
10 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
365
253
116
0
10
20
30
40
50
60
70
80
90
100
<00,
02-0
,04
0,06
-0,0
80,
1-0,
120,
14-0
,16
0,18
-0,2
0,22
-0,2
40,
26-0
,28
0,3-
0,32
0,34
-0,3
60,
38-0
,40,
42-0
,44
0,46
-0,4
80,
5-0,
520,
54-0
,56
0,58
-0,6
0,62
-0,6
40,
66-0
,68
0,7-
0,72
0,74
-0,7
60,
78-0
,80,
82-0
,84
0,86
-0,8
80,
9-0,
920,
94-0
,96
0,98
-1
B - 24h AM (µT)
num
ber o
f chi
ldre
np
0,4 µT
Mean exposures over 24 h – arithmetic mean (AM)
1.220.060.030.01Quartile in µT
99755025Proportion (%)
Bmean = 0.09µT
30 children with AM > 0.4 µT (3.1%)
11 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Mean exposures over 24 h – geometric mean
Bmean = 0.02µT
2 children with GM > 0.4µT
0.200.020.010.00Quartile in µT
99755025Proportion (%)
689
140
0
10
20
30
40
50
60
70
80
90
100
<0
0,02
-0,0
4
0,06
-0,0
8
0,1-
0,12
0,14
-0,1
6
0,18
-0,2
0,22
-0,2
4
0,26
-0,2
8
0,3-
0,32
0,34
-0,3
6
0,38
-0,4
0,42
-0,4
4
0,46
-0,4
8
0,5-
0,52
0,54
-0,5
6
0,58
-0,6
0,62
-0,6
4
0,66
-0,6
8
0,7-
0,72
0,74
-0,7
6
>0,7
8
B - 24h GM (µT)
num
ber o
f chi
ldre
n
NB id-expers
0,4 µT
12 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Exposures of type “clock-radio”
Proportion of children with AM > 0.4µT higher than in literatureSearch for explanations for these high exposures
Signal high during the night1/3 of harmonicsCorrespond to the field measured close to a clock-radio
Additional investigations High variation of B from a clock-radio to anotherSource = transformerB decreases very quickly with the distance (negligible at 50 cm)
Are these measurements representative of the personal exposure?Respect a distance of 50 cm between the EMDEX and any electric appliances during the nightQuestion asked in the questionnaire in order to checkThe measurements over 24h overestimate the exposure
Distinguish the exposure over 24h and the exposure outside sleepperiod
13 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Mean exposures outside sleep period
AM = 0.05 µT11 children with AM > 0.4 µT (1.1%)
GM = 0.02 µT
14 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
The sources of high mean exposures
Example of children, AM over 24h24 cases of EMDEX put very close to clock-radios during the night2 cases of EMDEX put close to electric appliance with a transformer during the day1 case of EMDEX put close to unknown electric appliance at home during the night and the day1 case of EMDEX put close to electric appliance with a transformer at school1 case on AC electrified train network close to the home and the school1 case of overhead low voltage line close to the home
Example of children, AM outside period of sleep5 cases of EMDEX put close to electric appliance with a transformer during the day1 case of EMDEX put close to unknown electric appliance at home the day1 case of EMDEX put close to electric appliance with a transformer at school1 case on AC electrified train network close to the home and the school1 case of overhead low voltage line close to the home1 case of electric network close to the school (to be confirmed)1 case of EMDEX put close to electric cable on the floor of a car
15 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Comparison of exposures (1/2)
Children / adultsChildren are less exposed than adults (over 24h and outside sleep period)
Home / outsideChildren are more exposed at home than outside (over 24h and outside sleep period)
Day / nightAt home, children are more exposed during the day than during the night
RegionChildren living in Ile-de-France are more exposed than in the other regions (over 24hand outside sleep period)
16 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Comparison of exposures (2/2)
Results electric networksMean exposures (at home and over 24h) are higher for children living close to electric networks than for those living far away from these networks
Mean exposures (at home and over 24h) are not different for children living close to high voltage networks and for those living close to 50 Hz train networks
Calculation to be done for distribution network (low voltage and 20 kV)
17 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Characterisation of exposure
Example of children (distribution network and type of train alimentation not taken into account yet)
---Heating energy = electric
-x--Time spent at school
+Time spent watching TV
17,2%
+
+
+
+
+
+
x
AM
Children over 24h
27,2%
x
x
+
+
+
+
+
+
x
GM
Name of the variable Children outside period of sleep
AM GM
Density of population of the department x x
To have put the EMDEX close to a clock-radio
Home close to high voltage overhead power lines + +
Home close to electric train networks + +
Population of the city (> 2 000 inhabitants) + +
Age x
To live in a building + +
Time spent on computer + +
Time spent in shopping centre x
Time spent in train transports x x
Time spent in non electric transports x
Level of explained variance 17,7% 29,8%
18 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Repartition to children around electric networks
Number of children living “close to electric networks”
45MV/LV substation524Underground cable LV
13MV/LV substation in building
331Underground cable 20 kV10Underground cable 63 to 150 kV11Underground cable 225 kV4150 Hz train network371Overhead line LV24Overhead line 20 kV11Overhead line 63 to 150 kV7Overhead line 225 kV4Overhead line 400 kV
Number of childrenType of network
Total of children = 977
19 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
How much is the distance indicator conservative?
Arbitral classification of subjects within the corridors as “exposed”to magnetic fields generated by electric networksBut the width of the corridors is overestimated
Do the magnetic field measurements of the “exposed” subjects show the influence of electric networks or not?24h variation of magnetic field generated by electric networks is quite characteristic
Visual check
20 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Example of a child – source = power line
Signal with little noise and proportional to a load curve of a power line
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
Tue 12:00 AM Tue 12:00 PM
Mag
netic
Fie
ld (u
T)
Time
Broadband ResultantHarmonic ResultantFundamental Resultant
May/28/200707:08:52 PM
May/29/200708:26:58 PM
21 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
12:00 AM 06:00 AM 12:00 PM 06:00 PM
Mag
netic
Fie
ld (u
T)
Time
Broadband ResultantHarmonic ResultantFundamental Resultant
Jan/09/200806:06:25 PM
Jan/10/200807:54:01 PM
Example of a child – source = middle voltage underground cable ?
Signal with a trend of a load curve during the night, but noisy
22 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
0
1
2
3
4
5
Thu 12:00 PM Fri 12:00 AM
Mag
netic
Fie
ld (u
T)
Time
Broadband ResultantHarmonic ResultantFundamental Resultant
May/03/200708:19:49 AM
May/04/200709:35:52 AM
Example of a child – source = train network
Very noisy signal, proportional to traffic and with a ratio of harmonic above zero
23 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Measurements with influence of an electric network
Note that:B measured = summation of all sources Signal is often at the limit of ground noise and non specificTaking into account the floor of buildings, the numbers for underground networks and substation in buildings would decrease
13
45
524
331
10
11
41
371
24
11
7
4
Number of children living close to
electric network
3
9
75
60
5
4
12
53
1
3
6
4
Number of children with influence of electric network
MV/LV substation
Underground cable LV
MV/LV substation in building
Underground cable 20 kV
Underground cable 63 to 150 kV
Underground cable 225 kV
50 Hz train network
Overhead line LV
Overhead line 20 kV
Overhead line 63 to 150 kV
Overhead line 225 kV
Overhead line 400 kV
Type of network
24 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Conclusion (1/2)
Objective to have a database of 1000 children and 1000 adults attained1st study of personal exposure of a population at the scale of a country3.1% of children have observed a AM > 0.4 µT
Main sources = clock-radio The real exposure of the person was overestimated
Outside period of sleep, 1.1% of children have observed a AM > 0.4 µT
More coherent with literature
Children are less exposed than adultsThe analysis of the mean exposures has shown that the retained variables do not allow alone to characterise these means
25 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Conclusion (2/2)
Factors of exposure were identifiedThese factors depends on the population considered (adults or children), the type of mean (arithmetic or geometric), and the scenario (over 24h or outside period of sleep)
Qualitative analysis of electric networks data show that:The part of the population whose exposure to 50 Hz magnetic field is influenced by high voltage power lines is smallThe criteria of distance chosen in this study is maximizing and thus overestimates logically the number of people whose exposure to 50 Hz magnetic field is influenced by electric networksIt is not conclusive that underground electric networks are really the source of exposure seen in some measurements
26 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
Perspectives
Continue the analysis include electric distribution network
Improvement of the characterization of mean exposures (variance explained) ?
Improve data on train networkTake into account the frequency
Other possible use of these data validation of physical models used to estimate magnetic fields
27 - I. Magne - NIR and Children's Health, Ljubljana - 18 May 2011
THANK YOU FOR YOUR ATTENTION !
This study was funded by the Ministry of Health and Solidarities and conducted by Supélec, with the technical collaboration and the financial support of EDF and RTE.