SAR procedures: absorption rate (IEC 62209, EN 50360)

Author Hugues Orgitello Last updated 30 August 2026 ~18 min read

Guide · RF exposure and SAR

Specific Absorption Rate, or SAR, is the reference quantity to characterise the radiofrequency power absorbed by biological tissue when a transmitter is used in contact with the body or in its immediate proximity. All phones, smart watches, wireless earbuds, smart glasses, audio headsets and body-worn medical devices are subject to it. The measurement method is horizontal, defined by the IEC 62209 series, and taken up by the European Union through EN 50360 and EN 50566, and by the FCC through OET Bulletin 65. This page lays out the regulatory limits on both sides of the Atlantic, the measurement standards, the test hardware (SAM phantom, robotic platforms), the switch to power density for mmWave bands, and the most frequent pitfalls in dossier preparation.

What SAR measures

SAR (Specific Absorption Rate) is defined as the electromagnetic power absorbed per unit mass of tissue, expressed in watts per kilogram. The mathematical definition is SAR = sigma * E^2 / rho, where sigma is the electric conductivity of tissue in siemens per metre, E the modulus of the internal electric field in volts per metre, and rho the tissue mass density in kilograms per cubic metre.

The quantity is local: it varies point by point in the biological volume. Regulations therefore impose two types of averaging before comparison to a limit.

• Spatial averaging. SAR is integrated over a cubic tissue volume of mass 1 g or 10 g depending on the regime. The FCC uses 1 g, the European Union 10 g.
• Time averaging. Integration is over a 6 minute window for controlled exposure and 30 minutes for the general public, per the FCC rules at 47 CFR 1.1310 and OET Bulletin 65. Under ICNIRP 2020 the window follows the quantity rather than the population: 6 minutes for local SAR and local absorbed power density, 30 minutes for whole-body average SAR.

This double averaging explains why the same product can show very different SAR values across reports: a raw, instantaneous and local SAR value has no regulatory meaning. Only the value averaged per the protocol of the standard cited in the dossier counts.

Why averaging mass changes everything

The big US-EU split is not just in the numeric value (1.6 vs 2.0 W/kg) but in the mass over which the average is taken. Averaging over 1 g (FCC) captures a stronger local concentration than averaging over 10 g (EU). At the same peak SAR, the value averaged over 1 g is structurally higher than the value averaged over 10 g. Practical outcome: a product that comfortably clears 2 W/kg over 10 g in the EU is not guaranteed to pass 1.6 W/kg over 1 g in the United States, and the analysis must be redone for each configuration.

Two distinct regulatory regimes

SAR is framed by two legal architectures that share the same scientific foundations (ICNIRP guidelines and IEEE C95.1) but with different regulatory transpositions.

EU side: RED Article 3.1(a)

For radio equipment placed on the European market, the health protection requirement rests on Article 3.1(a) of the RED Directive 2014/53/EU, which demands protection of human health and safety. The operational translation goes through:

• Council Recommendation 1999/519/EC, which fixes the numerical exposure limits for the general public, aligned with ICNIRP 1998 and then ICNIRP 2020.
• The harmonised standards EN 50360 for handheld phones used near the head, EN 50566 for other body-worn equipment, and the EN 62209 series (CENELEC transposition of IEC 62209) for the measurement procedure.

See RED pillar for the directive framework and RED tests for the detail of health tests.

US side: 47 CFR plus OET Bulletin 65

The FCC handles SAR in several places:

• 47 CFR 1.1307 identifies the categories of transmitters subject to RF exposure evaluation.
• 47 CFR 1.1310 fixes the numerical limits (Maximum Permissible Exposure and SAR), and 47 CFR 2.1093 details the procedure for portable devices.
• OET Bulletin 65 (Federal Communications Commission, Office of Engineering and Technology) provides the historical technical application guidance; Supplement C-0101 is discontinued and its content now sits in the KDB publications, cross-referenced in KDB 447498 D03.
• KDB publications (Knowledge Database) clarify special cases: KDB 447498 (general RF exposure guidance, including modules), KDB 865664 (SAR measurement up to 6 GHz), KDB 248227 (SAR for 802.11 transmitters), KDB 941225 (SAR for 3G and LTE devices, hotspot mode and UMPC mini-tablets), KDB 690783 (SAR listings on grants).

See FCC pillar and FCC tests for the US framework.

Limits compared

Quantity	European Union (1999/519/EC, ICNIRP)	United States (47 CFR 1.1310)
Localised head/trunk SAR	2.0 W/kg averaged over 10 g (6 min)	1.6 W/kg averaged over 1 g (30 min)
Localised limb SAR	4.0 W/kg averaged over 10 g (6 min)	4.0 W/kg averaged over 10 g (30 min)
Whole-body SAR	0.08 W/kg averaged over total mass	0.08 W/kg averaged over total mass
Frequency range	100 kHz to 6 GHz (SAR)	100 kHz to 6 GHz (SAR)
Above 6 GHz	Power density, ICNIRP 2020	Power density, KDB 447498 guidance + KDB inquiry

Reading: while whole-body and limb limits converge, the critical difference lies in localised head/trunk SAR. In the United States, the 1 g averaging mass yields by construction higher values at the same incident field, and the numerical limit is lower. The margin ratio between the two regimes depends on the spatial distribution of the field and cannot be predicted by a simple scalar factor.

§ § §

The IEC 62209 series

IEC 62209 is the cornerstone of SAR metrology. Published by IEC under TC 106 (electromagnetic fields in the human environment), it is transposed in Europe as EN 62209 by CENELEC. It provides the practical procedure: calibration protocol, phantom geometry, robotic control, internal field post-processing.

The four parts of the series

Part	Scope	Frequency range	Typical use case
IEC 62209-1	Handheld phones used near the head	300 MHz to 6 GHz	Smartphone held to the ear, DECT phone
IEC 62209-2	Body-worn devices	30 MHz to 6 GHz	Watch, medical sensor, hotspot, glasses
IEC 62209-3	Vector measurement-based (fast-SAR) systems	600 MHz to 6 GHz	5G smartphone with simultaneous cellular + Wi-Fi + BLE
IEC 62209-4	Reduced-scope test methods	30 MHz to 6 GHz	Product variants, validated numerical simulation

IEC 62209-1 covers the historical case of the phone held to the ear with the SAM phantom. IEC 62209-2 covers the body-worn case with a flat phantom. Part 3, published in 2019, formalises methods for products transmitting simultaneously over multiple technologies: this is the default situation on any modern smartphone. IEC 62209-4, more recent, proposes methods to reduce the measurement scope when a numerical simulation (typically FDTD) experimentally validated on a reference configuration is available.

Coverage by European standards

The harmonised EN standards reproduce the IEC content, sometimes with a publication delay or RED-specific annexes. The current correspondence is as follows.

EN standard	Equivalent IEC reference	OJEU RED status
EN 50360	Phone-specific methods, older	Listed, undergoing evolution
EN 50566	Body-worn specific methods	Listed
EN 62209-1	IEC 62209-1	Listed
EN 62209-2	IEC 62209-2	Listed
EN 62209-3	IEC 62209-3	Being integrated

See RED standards for the application rules of harmonised standards and their life cycle.

Test hardware: phantom, liquid, robot

A SAR campaign runs in a very specific environment, not easily transferable from a standard EMC lab.

The SAM phantom and the flat phantom

The SAM (Specific Anthropomorphic Mannequin) is defined in IEEE 1528 and referenced by IEC 62209-1. It is a low-loss dielectric plastic shell reproducing the head geometry of the adult male 90th percentile, with two factory-set ear positions. The SAM is filled with a tissue-equivalent liquid whose relative permittivity and conductivity are adjusted at the test frequency to mimic the average electromagnetic properties of the human brain. The liquid composition changes at each band: a typical water / sugar / salt / surfactant blend for 900 MHz is not suitable for 5 GHz.

For body tests, the SAM is replaced by a flat phantom: a flat-bottomed plastic parallelepiped filled with the same kind of liquid. The product is placed against the flat face at the separation distance declared by the manufacturer.

Dosimetric probe and robot

The internal electric field measurement uses a miniature dosimetric probe, typically a three-axis sensor with individually calibrated Schottky diodes. The probe is carried by a six-axis robotic arm that scans a predefined volume below the phantom skin. The de facto market standard is the DASY platform from SPEAG (Schmid & Partner Engineering AG, Zurich), present in most accredited SAR laboratories worldwide. Alternatives exist (cSAR3D also from SPEAG, Schmid systems, IndexSAR), but DASY still dominates the methodological literature.

Post-processing

The robotic scan produces a 3D map of the field modulus. The SAR software (DASY6 or equivalent) numerically integrates sigma * E^2 / rho over the required volume (1 g or 10 g), with the cubic averaging algorithm defined by IEEE C95.3-2002. The retained value is the maximum over the whole scanned volume for the configuration tested. Each configuration generates a map, an averaged value, and an archived calibration trace.

mmWave and the switch to power density

Above roughly 6 GHz, the penetration depth of the electromagnetic field into tissue drops sharply. At 28 GHz typically, the skin depth in human skin falls around 0.5 mm, and the absorbed energy concentrates in a superficial layer where the notion of SAR averaged over 1 g or 10 g loses physical meaning: integration would be over a volume mostly empty of absorption. Standards bodies therefore switched to a replacement quantity, the surface power density expressed in W/m^2.

ICNIRP 2020 and IEEE C95.1-2019

The ICNIRP 2020 guidelines and the IEEE C95.1-2019 update introduce time-averaged power density (TPD) as the reference metric above 6 GHz. Averaging is over a 6 minute window (controlled exposure) or 30 minutes (general public) and over an elementary surface of 4 cm^2 or 1 cm^2 depending on frequency.

Typical ICNIRP 2020 limits for the general public above 6 GHz are around 20 W/m^2 over 4 cm^2, with step relaxations up to 200 W/m^2 by frequency and scenario.

FCC and 5G FR2 mmWave bands

The FCC handles the mmWave bands (24 GHz, 28 GHz, 37 GHz and 39 GHz, used by 5G FR2 and certain backhaul links) as power density under 47 CFR 1.1310 and 2.1093, with KDB 447498 guidance and pre-approval KDB inquiries. The procedure combines:

• A power density measurement with a directional probe.
• An antenna pattern characterisation for beam-steering antennas (beamforming).
• A statistical exposure evaluation accounting for beam activation duration.

IEC 62232 provides the IEC normative framework for these measurements, complemented by ITU-T K-series guides for base stations.

Summary table: SAR vs power density

Frequency range	Quantity	Main standard	Method	Typical general-public limit
100 kHz to 10 MHz	Induced current density + SAR	ICNIRP 2020	Induced current measurement	Frequency-dependent
10 MHz to 6 GHz	Localised + whole-body SAR	IEC 62209-1/-2/-3	Phantom + robotic probe	2 W/kg / 10 g (EU), 1.6 W/kg / 1 g (US)
6 GHz to 300 GHz	Power density	IEC 62232, FCC RF exposure rules (KDB inquiry)	Directional probe, beam scan	10 W/m^2 over 4 cm^2 (ICNIRP 2020)

Test conditions and product configuration

The test procedure imposes a worst-case product configuration: this is the main challenge in dossier preparation.

Full power, multi-channel, multi-position

The product must be evaluated with:

• Maximum power on each band, typically driven through a test firmware or chipset-specific AT command.
• All radios active in credible use configurations: cellular, Wi-Fi 2.4 GHz, Wi-Fi 5 GHz, Wi-Fi 6 GHz, BLE simultaneously where relevant.
• Several channels per band: typically low, mid, high channels.
• Several positions against the phantom: head left, head right, cheek contact, ear contact (the IEEE 1528 cheek and tilt), body front, body back, several orientations.

For a multi-band smartphone, the number of configurations to test can exceed 100. Numerical simulation (experimentally validated FDTD) and reduced test scope (IEC 62209-4) techniques are then used to target worst cases and limit the robotic scan.

Liquid temperature

The tissue-equivalent liquid must be temperature-stabilised (typically 20 to 22 degrees Celsius), as its permittivity and conductivity vary appreciably with temperature. A drift of more than 2 degrees during the session invalidates the measurements. The lab calibrates the liquid at the start of each day and verifies it with a reference probe.

Separation distance

The distance between product and phantom face is a parameter declared by the manufacturer and recorded in the user notice. Common values range from 0 mm (direct contact) for a watch or medical patch, 5 mm for a smartphone, 10 to 15 mm for a belt-worn device. Under-declaring this distance to ease test passage is a classic pitfall: the notice must be consistent, and any marketing copy suggesting closer use creates a non-compliance risk.

KDB 941225 and hotspot-mode SAR

The US procedures for cellular technologies and for hotspot mode are not in OET Bulletin 65: they sit in KDB publication 941225, last updated 9 April 2024. Its numbering has moved, so the version matters: the publication's own note records that attachment D01 v03r01 consolidates the previous versions of D01, D02, D03 and D04 into a single document. Five attachments are current.

Attachment	Scope
941225 D01 v03r01	3G SAR procedures, 3GPP and 3GPP2 (1xRTT, EvDo, UMTS/WCDMA, GSM/GPRS)
941225 D05 v02r05	Release 8 LTE devices, FDD and TDD
941225 D05A v01r02	KDB inquiry sheet for LTE Release 10
941225 D06 v02r01	Portable devices with wireless router (hotspot) capabilities
941225 D07 v01r02	UMPC mini-tablet devices, diagonal 20 cm or less

Where hotspot mode is measured

D06 covers the device that routes traffic from nearby Wi-Fi clients through its own WWAN connection, as distinct from an access point routing over a wired link. Two rules fix the geometry, and they answer different questions.

• Form factor sets the separation distance. Where the overall length and width exceed 9 cm by 5 cm, hotspot-mode SAR is measured at 10 mm; for smaller devices 5 mm or less is required. Where this is unclear, D06 directs the applicant to a KDB inquiry rather than to engineering judgement.
• A 25 mm antenna rule sets the surfaces. SAR is measured on all edges and surfaces with a transmitting antenna located within 25 mm of that surface or edge. A face with no antenna within 25 mm never goes on the test list.

Because test distance plus 1-g measurement also covers hand-held exposure, a separate 10-g extremity evaluation is not necessary. With hotspot mode active during a call, D06 warns, up to four transmitters may transmit simultaneously, and hotspot traffic is typically transparent and unattended: the lab cannot assume the user has put the device down. Departing from these procedures is not a free choice either, since D06 requires a PAG (pre-approval guidance, KDB 388624) when they are not followed.

Deciding which configurations are actually required

The FCC never asks for the full combinatorial matrix. KDB 941225 hands the arithmetic to KDB 447498 D01, where the decision runs as a cascade, applied separately to each band, exposure condition and test position.

1. Standalone exclusion, section 4.3.1: no SAR evaluation when [(max. power of channel, including tune-up tolerance, mW) / (min. test separation distance, mm)] · √f(GHz) ≤ 3.0 for 1-g SAR, over 100 MHz to 6 GHz at separations of 50 mm or less. For hotspot mode the distance here is the phantom-to-device distance, 10 mm or 5 mm, never the 25 mm antenna-to-surface distance.
2. Channel reduction, section 4.4.1: the other channels of a band are skipped when the reported 1-g SAR of the mid-band or highest power channel is ≤ 0.8 W/kg for a band up to 100 MHz wide, ≤ 0.6 W/kg from 100 to 200 MHz, ≤ 0.4 W/kg at 200 MHz and above.
3. Simultaneous transmission, section 4.3.2: when the sum of the reported standalone 1-g SAR of all simultaneously transmitting antennas stays within the limit, the configuration is excluded outright. When the sum exceeds it, antennas are taken pair by pair and the SAR to peak location separation ratio, SPLSR, is computed as (SAR1 + SAR2)^1.5 / Ri, Ri being the distance in millimetres between the two peak SAR locations. The configuration is excluded when the ratio is ≤ 0.04 for every pair, or ≤ 0.10 where 10-g SAR applies. Pairs that fail are measured.

One threshold governs entry to the matrix: simultaneous transmission SAR is addressed when the maximum duration of overlapping transmissions, network handovers included, exceeds 30 seconds.

Two cautions. Where a battery-operated router needs an approved external dongle or card for hotspot mode, D06 assigns that transmitter a reported SAR of 1.6 W/kg, the whole 1-g limit: the sum test is withdrawn and the pair ratio must be used. And since 13 August 2022 either 447498 D01 v06 or the newer 447498 D04 Interim v01 may be applied, each in full and with no mixing; the ratio and channel thresholds are identical in both, the first-stage arithmetic is not. Treat that pairing as dated rather than permanent: the KDB 447498 listing states that after the transition period all applications must use new procedures to be published as 447498 D01 v07, replacing v06, and that the FCC will give notice of the transition period when v07 becomes available. Check the publication listing before you fix a test plan, because the choice of guidance document is the first thing a TCB looks at.

The contribution summation described in the next section is the same arithmetic seen from the system side: the first test of this cascade, applied to the finished product.

Simultaneous transmission scenarios

All smartphones and most modern wearables transmit simultaneously over several technologies: cellular + Wi-Fi + BLE, or less frequently cellular + cellular (Dual Connectivity 4G+5G).

Conservative approach: contribution summation

The simplest method measures the SAR of each radio individually and sums the contributions:

SAR_total = SAR_cellular + SAR_WiFi + SAR_BLE

This arithmetic sum is conservative: it assumes that the spatial worst cases superpose exactly, which is not the case in reality. It eases the compliance demonstration but penalises product margin.

Realistic approach: composite measurement

IEC 62209-3 and KDB 447498 allow a composite measurement by simultaneously driving the transmitters at their worst-case configuration and directly recording the resulting SAR map. The protocol requires specific equipment (synchronous radio control), a real-simultaneity validation procedure, and a statistical post-processing. The margin gain can be significant (typically 1 to 3 dB) but the cost and campaign duration grow.

Classic pitfalls in SAR dossier preparation

First-pass rejections in SAR labs come overwhelmingly from five recurring causes.

1. Inconsistent or undeclared separation distance. The manufacturer delivers the product without having formalised the separation distance to test. The lab applies a conservative default (typically 0 mm for a wearable, 5 mm for a smartphone), and the SAR value exceeds the limit. The distance must be fixed at specification, justified by realistic use, and printed in the user notice. See RED pitfalls for pitfalls tied to user documentation.

2. Wrong phantom. Testing a body-worn wearable with the SAM (head) phantom instead of the flat (body) phantom is an obvious error, but the reverse is more subtle: a hybrid product (smart glasses touching the head and temple at the same time as being worn in front) may require a double head + body evaluation with two successive phantoms.

3. Sub-6 GHz / mmWave confusion. For a 5G smartphone combining FR1 (sub-6 band) and FR2 (mmWave), SAR is needed on FR1 and power density on FR2, in the same campaign but with two distinct protocols. Confusing the two or testing only one is grounds for immediate FCC TCB rejection. The applicable FCC KDB guidance (447498 and the mmWave inquiry responses) must be read before the test phase, not discovered at the report stage.

4. Multi-radio configuration under-evaluation. The dossier covers only standalone cellular and omits the cellular + Wi-Fi + BLE combination. The TCB asks for a complementary evaluation or an aggregate SAR calculation under IEC 62209-3. Result: 1 to 3 weeks of additional delay and a second lab session.

5. Expired liquid calibration. The tissue-equivalent liquid must be re-titrated periodically (typically every 1 to 3 months depending on lab). A report produced with an out-of-validity liquid is rejected by the TCB. The check is administrative but often missed at dossier audit. See FCC pitfalls for dossier audit pitfalls.

Relationship with HAC compatibility

Hearing Aid Compatibility (HAC), governed in the United States by 47 CFR 20.19, requires that mobile phones be tested for their near-field emission in the audio band (around 1 kHz typically) and their coupling with hearing aids. Although distinct from SAR (phantoms, bands and criteria differ), the HAC campaign is often organised in parallel with SAR because:

• The product is already on the bench in worst-case configuration.
• SAR-equipped labs are generally also HAC-equipped.
• The overall campaign schedule is thus optimised.

HAC criteria are M3 or M4 for audio magnetic coupling and T3 or T4 for T-Coil coupling, with M4/T4 stricter than M3/T3. See the glossary for full term definitions: SAR, HAC, MPE.

Campaign timeline and cost

To plan budget and lead time to market, the orders of magnitude observed for a campaign in a European lab accredited ISO/IEC 17025 and FCC-recognised are as follows.

Product type	Typical configurations	Campaign duration	Indicative cost
Single-band BLE wearable	4 to 8	3 to 5 days	EUR 4 000 to 8 000
4G LTE multi-band smartphone	20 to 50	1 to 2 weeks	EUR 12 000 to 25 000
5G smartphone FR1 + FR2 (mmWave)	50 to 100+	3 to 4 weeks	EUR 25 000 to 50 000
Body-worn medical device	5 to 15	1 week	EUR 6 000 to 12 000
Smart glasses (head + body)	10 to 25	1 to 2 weeks	EUR 10 000 to 18 000

Cost depends mainly on the number of configurations to evaluate, and secondarily on robot mobilisation for the longest scans (mmWave under active beamforming can multiply durations). A dual EU + US campaign in an FCC-MRA lab can produce a single report usable on both sides, provided that the dual coverage was specified at quotation. See Certification costs for the overall budget framework.

Key takeaways

• SAR measures RF power absorbed per unit tissue mass. EU limits: 2 W/kg over 10 g; US limits: 1.6 W/kg over 1 g. Averaging mass matters as much as the numerical value.
• Metrology rests on IEC 62209. Four parts cover phones (1), body-worn (2), multi-source (3) and reduced-scope derived methods (4).
• Phantoms are specific. SAM for the head, flat for the body, filled with tissue-equivalent liquid calibrated per band.
• Above 6 GHz, SAR is replaced by power density (TPD), governed by IEC 62232 and the FCC RF exposure framework for 5G FR2 mmWave.
• The declared separation distance is a critical dossier parameter: set it upfront and verify it across all user documentation.
• A multi-band campaign takes 2 to 4 weeks for a complete product, more with active beamforming or simultaneous multi-radio transmission.

For practical implementation on the EU side, see RED tests. For the US side, see FCC tests. For term definitions, see the Glossary.

See also

• IEC 61000-4-3: radiated RF field immunity
• IEC 61000-4-6: conducted RF immunity
• IEC 61000-4-8: power-frequency magnetic field immunity
• ESD per IEC 61000-4-2: method and test levels

Working on antenna specification and product design under SAR constraint for a real product? AESTECHNO can audit your design and certification plan. Get in touch →

Sources & references

1. IEC 62209-1:2016, Measurement procedure for SAR for handheld mobile devices (300 MHz to 6 GHz) , IEC webstore.iec.ch/en/iec-search/result?q=IEC%2062209-1
2. IEC 62209-2:2010, Measurement procedure for SAR for body-worn devices , IEC webstore.iec.ch/publication/6590
3. IEC 62209-3:2019, Vector measurement-based systems (600 MHz to 6 GHz) , IEC webstore.iec.ch/en/iec-search/result?q=IEC%2062209-3
4. FCC OET Bulletin 65, evaluation of RF field exposure , FCC OET www.fcc.gov/general/oet-bulletins-line
5. 47 CFR section 1.1310, radiofrequency radiation exposure limits , FCC / eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-1/subpart-I/section-1.1310
6. Council Recommendation 1999/519/EC on the limitation of exposure of the general public to electromagnetic fields , EUR-Lex eur-lex.europa.eu/eli/reco/1999/519/oj
7. ICNIRP 2020 Guidelines, exposure to electromagnetic fields 100 kHz to 300 GHz , ICNIRP www.icnirp.org/en/publications/article/rf-guidelines-2020.html
8. FCC KDB Publication 941225, SAR test procedures for 3GPP and 3GPP2 devices (D01 3G, D05 LTE, D06 hotspot mode, D07 UMPC mini-tablet) , FCC OET apps.fcc.gov/oetcf/kdb/forms/FTSSearchResultPage.cfm?switch=P&id=26930
9. FCC KDB Publication 447498, RF exposure procedures for mobile and portable devices (D01 v06, D04 Interim v01) , FCC OET apps.fcc.gov/oetcf/kdb/forms/FTSSearchResultPage.cfm?switch=P&id=20676