Safety limits - Details

  • See EUROPAEM EMF Guidelines 2016 for the latest biological safety limits
     - Radio Frequency radiation safety limits  (uW/m2): Table 3
               - conversion of Radio-Frequency radiation units from uW/m2 to V/m: Table 4
     - VLF magnetic fields limits (nT): Table 5
     - VLF electric fields limits (V/m): Table 6
     - ELF magnetic fields limits (nT): Table 1
     - ELF electric fields limits (V/m): Table 2

  • In comparison, ICNIRP's limits are only for heating and not for established long-term and low-level biological effects, although the European Directive of 2013 does specifically include non-thermal electrosensitivity symptoms such as vertigo and nausea for low frequencies.

Guidelines like EUROPAEM 2016 now specify:

  • intensity of exposure
  • length of exposure - e.g. 20 out of 24 hours should be under the specified limit,
  • location or variability of exposure - e.g. day time and thus variable, or night time in one place
  • person exposed - e.g. 'sensitive' to the radiation, or 'general population'
  • type of radiation - e.g. by amplitude, frequency, modulation or pulse.

Power Density

microWatt / meter squared

naturealtered EEGsome
ES symptoms
safety limit (BioInitiative, 2012)
safety limit (ICNIRP, 1998)

(1800 MHz)

(900 MHz)

(general population)

​(children, sensitives)




Power Density: radio frequency

microWatt / meter squared

>4 hours per day
20 out of 24 hours should be below these limits

Night-time exposureDay-time exposure

 WiFi    (10 Hz)
 GPRS  (8.3 Hz)
 DAB+  (10.4 Hz)


 DECT cordless phone

 3G   (UMTS)
 4G   (LTE)


 FM radio


Electric Fields

peak to peak

Volt / meter

ES symptoms

safety limits

(Council of Europe, 2011)

safety limit
(ICNIRP, 1998)



(medium term)



Magnetic Fields

time-varying, 50-60 Hz


human sensitivity
(Aurora Disturbance, solar flare)

human brain

entrainment: Schumann Resonance

biological safety limit
(BioInitiative 2007)
heating safety limit
(ICNIRP, 2008)





(rise/fall, at 0.0013)

Power frequency: 50 Hz - 2 kHz


>20 hours per week

Sensitive populationsNight-time exposureDay-time exposure
 nT maximum30010001000

Intermediate frequencies: 3 kHz - 3 MHz


dirty power, PLC, RFID, CFL

>4 hours per day

20 out of 24 hours should be below these limits

Sensitive populationsNight-time exposureDay-time exposure

Body currents


>4 hours per day

20 out of 24 hours should be below these limits

Sensitive populationsNight-time exposureDay-time exposure



Specific Energy Absorption Rate

Watts / kilogram

SAR applies only to acute heating,

averaged for 6 minutes for simulated tissue mass.

SAR is irrelevant to ES and biological and long-term effects.
See below for biological limits, such as DNA damage, to replace SAR.

biological damage thresholdbiological safety limit
(Seletun, 2010)
SAR heating safety limit
(whole body)
SAR heating safety limit
SAR heating safety limit


1.6 or 2.0




Safety Limits - Summary

Experts: zero exposure only

Three leading experts in the field of bioelectromagnetics have stated that
the only valid safety limit for man-made electromagnetic exposure is zero.
For the heating measure of power density this means:

                0 microWatts/cm2   =   0 milliWatts/cm2   =   0 microWatts/m2

This zero-only approach is the same for ionizing nuclear radiation,
where it is generally accepted that there is no safe exposure limit to man-made nuclear radiation. 

Therefore any acceptance of man-made radiation is inherently risky.
It is likely to cause adverse effects in some biological organisms even at the lowest detectable levels.

Dr George Carlo: ​
"Dr. Carlo confirms that cell damage is not dose dependant because any exposure level,
no matter how small, can trigger damage response by cell mechanisms."  
"Wherever you have information that is being transmitted wirelessly, 
you have the ability to trigger those protective responses and that is dangerous, there is no threshold."

              Assoc. Prof. N Cherry:

"He also found many epidemiological studies showing dose-response relationships
for cancer, cardiac, reproductive and neurological effects,
showing a safe level of zero exposure, consistent with EMF/EMR being genotoxic." 

Assoc. Prof. O Johansson: 
"A completely protective safety limit based on today’s information would be zero."
(For reference details, see below.) 

International  Electromagnetic  Biological  Safety  Limits


Field measuredTypeMetric

Natural  or

background levels

Safety Limit
1Magnetic Field

power lines

(50-60 Hz)

nanoTesla0.0002 nT10 nT
2Electric Field

radio waves

(>30 kHz)

Volts per meter0.00002 V/m0.006 V/m
3Electric Field

power lines

(50-60 Hz)

Volts per meter

0.0001 V/m1 V/m


power flux density


radio waves

(pulsed require lower limits)

microWatt per meter squared
0.000001 uW/m20.1 uW/m2


specific energy absorption rate (SAR)

radio wavesWatts per kilogram< 0.00002 W/kg0.0003 W/kg


      NB: Electric fields are measured peak to peak, not averaged.
             Biological safety limits are for long-term and cumulative exposures.

             Heating limits (e.g. ICNIRP) are set for only 6 minutes, averaged.

New metrics have recently been suggested as more accurate and relevant for measuring biological safety limits, such as assessing DNA damage, protein expression, fertility impairment, oxidative stress, tumor promotion, etc.

For sources of the International Safety limits, see below (Building Biology SBM-2008, BUND, EUROPAEM 2016, Seletun 2010 etc.).

These biological International Safety Limits are:

  • for all the general population, including children, the elderly, people with sickness, people sensitive to electromagnetic exposure, people with cancer or neurological or inflammatory illnesses,  etc.
  • for all schools, homes, work places and public areas, where adults and children with different levels of sensitivity spend long periods of time.
  • for all long-term and cumulative exposures, unlike the acute heating limits (e.g. ICNIRP) which are designed to prevent one degree of heating within six minutes.

A survey of personal exposure in Slovenia in 2016 found an average of 0.26 V/m, far above the international EUROPAEM 2016 safety limit of 0.006 V/m for sensitive people and people with EHS. (Gajšek P et al, IEEE Rad Anten Days Ind Ocean (RADIO), 2016)

Biological limits

  • BioInitiative Report: "A Rationale for Biologically-based Public Exposure Standards for Electromagnetic Fields (ELF and RF)" (2012) 
  • Building Biology: Guidelines (2008)
  • BUND (Bund Umwelt und Naturschutz Deutschland): in Grenzwerte für elektrische und magnetische Felder (2016)
  • Damvik M, Johansson O: "Health risk assessment of electromagnetic fields: a conflict between the precautionary principle and environmental medicine methodology" Rev Environ Health (2010) PMID: 21268445
  • EUROPAEM EMF Guidelines (2016; pdf)
  • Fragopoulou A et al.: "Scientific panel on electromagnetic field health risks: consensus points, recommendations, and rationales" Rev Environ Health (2010) PMID: 21268443 [Seletun, 2010]
  • Hardell L, Sage C: "Biological effects from electromagnetic field exposure and public exposure standards" Biomed Pharmacother. (2008) PMID: 18242044
  • Leszczynski D, Xu Z: "Mobile phone radiation health risk controversy: the reliability and sufficiency of science behind the safety standards" Health Res Policy Syst. (2010) PMID: 20205835
  • Powerwatch: "International Guidance Levels" (about 2008)
  • Salzburg: "Precautionary limits" (2002)

SAR heating limits should be replaced by non-linear biological limits,

e.g. DNA fragmentation

SAR heating limits (Specific Absorption Rate) are still used by the few remaining thermalists - scientists still claiming that the only effect of radiation is heating. This claim was disproved in the 1960s and the majority of scientists now accept non-thermal effects. Thus, of 30 scientists in 2011 at the World Health Organization's IARC, 28 voted to classify non-thermal radiation as a 2B cancer agent while only 2 still denied this because of the invalidated heating hypothesis. In fact many medical procedures now depend on non-thermal radiation, convincingly disproving the thermalists' claims. A heat rise of 1.0 C is regarded as dangerous by thermalists, despite the fact that even slight exertion by the human body quickly raises the body temperature by a similar or greater amount. Thermalists disregard all the established evidence for other changes, such as molecular, CNS, protein expression, blood perfusion, DNA fragmentation etc, all now regarded by the majority of scientists as more useful and valid indications of adverse effects.

  • SAR as a limit for EM exposure is irrelevant for ES sufferers. It was designed for heating effects and ES is not caused by heating.
  • SAR, measured by W/kg, was developed during the 1950s by Herman Schwan with other thermalists like Piersol and Li in the USA. Schwan was one of the thermalists most responsible for the invalid assumptions made by the USA during the 1950s in the adoption a purely thermalist viewpoint on the non-thermal effects of electromagnetic exposure.
  • The 4 W/kg limit, still used today, was formulated in the USA in 1982. In 1981 NCRP SC 39 proposed heating and SAR for determining an exposure limit, and ACGIH integrated SAR into their exposure standards in 1981, as did ANSI (C95.1-1982) in 1982.
  • This 4 W/kg heating limit was adopted by pro-wireless groups like the USA's IEEE C95 committee in 1992. In 1998 the pro-wireless ICNIRP also adopted a similar heating standard, but averaged over 10g of tissue instead of 1g used in the USA.
  • As usual with toxic substances, reductions were applied to allow a safety margin. The 4 W/kg threshold for SAR was reduced by 10 to 0.4 W/kg for occupational exposures. It was reduced by 50 to 0.08 W/kg for the general public.
  • In 1986 the protection for the general public was reduced to 1.6 W/kg, at 1/5th of the occupational limit of 8 W/kg (NCRP Report No.86 (1986) “Biological Effects and Exposure Criteria for Radio Frequency Electromagnetic Fields”, USA) This was based on the assumption that the general public's continuous exposure  is the maximum of 168 hours per week compared with 40 hours per week for occupational environments.
  • In comparison, in 1985 NIOSH proposed a lower limit of 2 W/kg and the EPA 1 W/kg. 
  • In 1996 the FCC (96-326) adopted 8 W/kg  or 1.6 W/kg for partial body exposure and 0.4 or 0.08 W/kg for whole body exposure, averaged over 6 minutes.
  • In 1998 the WHO's ICNIRP (Health Physics, 1998) adopted SAR averaged over 6 minutes and 10 g tissue at 0.08 W/kg for whole body exposure, 2 W/kg for localised SAR (head and trunk), and 4 W/kg for localised SAR (limbs).
  • SAR heating standards are based on tests carried out during the 1970s and 1980s. These were on animals, such as monkeys and rats, and not on humans. The threshold was determined as the disruption of food-motivated learned behavior in subject animals. This has nothing to do with many of the cognitive, neurological and cancer effects now established for non-thermal radiation exposure, meaning that the 4 W/kg SAR limit is worthless in protecting humans from many established adverse effects of electromagnetic radiation.
  • SAR heating limits are averaged over 6 minutes and therefore do not cover long-term exposure. They were designed for adult healthy males. They do not cover children, the elderly, the sick or those with ES.
  • SAR heating limits and similar heating limits do not take into account the multiple sources of radiation from the numerous cellphones, WiFi routers, cellphone towers and similar devices now around us all for much of the day and night (Ajay Malik: "Why the FCC's safety guidelines for Wi-Fi need to be re-evaluated" Network World, November 12 2015)

Most medical experts now recommend that SAR should be replaced with more accurate and relevant non-thermal, biological and long-term tests, such as DNA fragmentation, cell hydration or oogenesis.


  • Anon. "Specific Absorption Rate, or SAR – FCC Cell Phone Radiation Exposure Limits" (RF Safe)
  • Ayrapetyan S et al: “Cell hydration as a biomarker for estimation of biological effects of nonionizing radiation on cells and organisms” ScientificWorldJournal. (2014) PMID: 25587574.
  • ​Blank M et al: "Electromagnetic fields and health: DNA-based dosimetry" Electromagn Biol Med. (2012) PMID: 22676645
  • Margaritis LH et al: “Drosophila oogenesis as a bio-marker responding to EMF sources” Electromagn Biol Med. (2014) PMID: 23915130.
  • Panagopoulos DJ et al., "Evaluation of specific absorption rate as a dosimetric quantity for electromagnetic fields bioeffects" PLoS One. (2013) PMID: 23750202

The heating hypothesis assumes a linear or dose-response biological effect. In fact it has long been known that low-level non-thermal effects are non-linear and depend on the particular frequency, modulation and exposure duration or repetition.

  • Sagioglou NE et al: “Apoptotic cell death during Drosophila oogenesis is differentially increased by electromagnetic radiation depending on modulation, intensity and duration of exposure” Electromagn Biol Med. (2016) PMID: 25333897.

Most scientists now reject the IEEE, ICNIRP or FCC denial of established non-thermal effects. Thus since the 1980s in the USA leading scientists have rejected ANSI / IEEE claims that their thermal limits protect against known non-thermal effects:

  • "This adverse effect level for human beings, 4 W/kg, is the threshold for a specific biological effect, i.e., behavioral disruption (work stoppage) in nonhuman primates that is associated with an increase in body temperature. Work stoppage, the failure of a food-deprived animal to perform a learned task to gain a food reward, is interpreted to result from  thermal stress, caused by the absorption of RF energy, that is sufficiently severe to deter hungry animals from working for food.
  • Since the ANSI/IEEE hazard level is an SAR associated with an effect resulting from a known mechanism of interaction (RF heating) that is associated with an increase in body termperature (as is the NCRP hazard level), the ANSI/IEEE C95.1-1992 standard is based on a thermal effect of RF radiation and, by extension, is portective of effects arising from a thermal mechanism, but not from all possible mechanism. Therefore, the generalization that 1992 ANSI/IEEE guidelines protect human beings from harm by any mechanism is not justified."
  • "In contrast to the 1992 ANSI/IEEE standard, 1986 NCRP states that a response to RF radiation may have a "thermal basis, an athermal basis, or a combined basis," and that a "determination of which of these three classes of causation is operative in a given context rests upon appropriate experimentation and inference, not presumption." NCRP also claims that there is "no intent to define exposure criteria solely in terms of SAR," and that "consideration is also given to other factors where appropriate." These factors include, among others, possible modulation- and carrier- frefquency specific biological responses."    
    (EPA, Air Quality & Radiation:  "EPA Submission to FCC", 1993)

Cell Towers and Radiation Exposure Levels

Precautionary Principle

  • “Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation.”
    Principle 15 (UN Conference on Environment and Development, Rio de Janeiro, 1992)

Simplified History of Safety Limits

A major divergence in safety limits, of 1000 times, can be seen between heating values adopted by the USSR and the USA from 1935 to the present day.

The EUROPAEM limits (2016) diverge even more, by 10 million to 1 billion times. They adopt a more up-to-date approach in the light of recent science showing adverse effects from exposure at much lower levels, but which are still far above natural levels.

Leading experts state that, as for ionizing radiation, there is no safe level of man-made non-ionizing exposure above natural levels.

[Watts per meter squared is essentially a heating metric, as first used in 1935 to prevent heating.
Modern metrics should be based on perhaps DNA breaks or protein expression as a more reliable and accurate metric: see above.]

The figures below are simplified and general, since more recent standards have variations for some frequencies, pulse and amplification characteristics and exposure durations.


1935: perhaps the first RF safety limit, for physiotherapy rooms: 

              10 microWatts/cm2   =   0.01 milliWatts/cm2   =   100,000 microWatts/m2

This limit of 100,000 uW/m2 has remained the USSR and Russia's safety limit for workers (1955), and the general population (1968, 1984, 1996 and 2003), or 0.2 V/m.


1953: H Schwan proposed for the US Navy Dept.:

               10,000 microWatts/cm2   =   10 milliWatts/cm2   =   100,000,000 microWatts/m2

This limit of 100,000,000 uW/m2 remained the USA's safety limit for the general population (1966 ANSI, 1975, and IEEE 1991, 1998 ).  

Experts: zero exposure only

Three leading experts in the field of bioelectromagnetics have stated that the only valid safety limit for man-made electromagnetic exposure is zero.

                0 microWatts/cm2   =   0 milliWatts/cm2   =   0 microWatts/m2

This zero-only approach is the same for ionizing nuclear radiation, where it is generally accepted that there is no safe exposure limit to man-made nuclear radiation.

Therefore any acceptance of man-made radiation is inherently risky. It is likely to cause adverse effects in some biological organisms even at the lowest detectable levels.


2016: heating safety limit for sensitive populations: 
                0.000001 microWatts/cm2   =    0.00000001 milliWatts/cm2   =   0.1 microWatts/m2

           heating safety limit (daytime) for general population: 
                0.0001 microWatts/cm2   =   0.000001 milliWatts/cm2   =   10 microWatts/m2

Reduction to 5 V/m (2017)