Energy meter

01

Choose the accuracy class and its error limits

Annex V (MI-003) of the EU Measuring Instruments Directive 2014/32/EU covers active electrical energy meters for residential, commercial and light industrial use and requires the maker to declare class A, B or C. Its maximum permissible error table for single-phase meters, and polyphase meters on balanced load, gives 3.5% for class A, 2% for class B and 1% or 0.7% for class C at +5 to +30 °C depending on the load-current band, widening to 9%, 4% and 2% at −40 to −25 °C or +55 to +70 °C. Meters must meet these limits over a power factor range of at least 0.5 inductive to 0.8 capacitive. [1]

The directive says member states imposing residential measurement must accept any class A meter and may require class B for specified purposes, and for commercial or light industrial measurement must accept class B and may require class C. It sets critical change values for disturbances, for example 4%, 2% and 1% for voltage unbalance and 6%, 3% and 1.5% for DC and harmonics in the current circuit for classes A, B and C, and requires the meter to comply with electromagnetic environment E2. [1]

Editorial calculation: a class B meter's limit of 2% at room temperature means a meter registering 10,000 kWh a year could be up to 200 kWh high or low and still conform; the same meter operating at −25 to −10 °C or +40 to +55 °C is allowed 3.5%, or 350 kWh. [1]

Give the supplier the supply (single-phase, three-phase three- or four-wire), reference voltage and frequency, maximum current or CT ratio, direct or transformer connection, required accuracy class and legal approval (for example MID), import/export and reactive energy needs, tariffs, outputs (pulse, Modbus RTU, M-Bus, wireless), display, mounting (DIN rail or panel), operating temperature, sealing, and quantity.

02

Size the current range, register and outputs

MI-003 defines the currents a maker must declare: the starting current at which the meter begins to register, the minimum current above which it is within its error limits, the transition current above which the tighter limit for its class applies, and the maximum current; transformer-operated meters also declare a reference current. The register must have enough digits not to return to zero after 4,000 hours at maximum current, reference voltage and unity power factor, must not be resettable in use, and measured energy must stay readable for at least 4 months after loss of supply. The meter must not register with voltage applied and no current, and must display kilowatt-hours or megawatt-hours. [1]

Eastron's SDM630 is a three-phase, 100 A direct-connected, four-module DIN-rail meter, MID certified to modules B and D, that measures 1p2w, 3p3w and 3p4w supplies, is self-powered from the mains inputs, and comes with Modbus RTU and dual pulse outputs, M-Bus and pulse, pulse only, or multi-tariff options. It displays voltage, current, THD, frequency, power factor, demand, active, reactive and apparent power, and import, export and total active and reactive energy, and lists EN 50470-1/-3, EN 61326 and EN 61010 standards. [2]

Editorial calculation: at a 230 V reference voltage, a three-phase meter with a 100 A maximum current runs at about 3 × 230 × 100 = 69 kW at full load, so 4,000 hours is about 276,000 kWh — the register needs at least six whole-kWh digits to meet the MID rule. [1] [2]

03

Compare what suppliers offer

Suppliers linked to this page list energy meters; each card links to the page reviewed.

Listings include DIN-rail and panel energy meters, prepayment and smart three-phase meters, DC energy meters, multi-channel meters and LoRaWAN wireless meters; ask each for the accuracy class and any legal approval, current range, connection type, outputs and protocol documentation.

04

Ask for system accuracy when the meter uses external CTs

Continental Control Systems' application note AN-136 warns that a submeter's accuracy claim may exclude its external current transformers, and that a 0.2% meter with 0.3% CTs may not result in 0.5% accuracy. The note explains that the U.S. CT standard IEEE C57.13 only specifies accuracy down to 10% of rated current and allows twice the error there, so a class 0.3 CT could have a 0.3% error at 100% of rated current, 0.6% at 10% and an arbitrarily large error at 1%. IEC 61869-2 and IEC 60044-1 classes 1, 0.5, 0.2 and 0.1 stop at 5% of rated current and allow errors 3.0 to 4.0 times larger there; classes 0.2S and 0.5S reach 1% but allow 3.0 to 3.75 times larger errors. The ANSI C12.1 and C12.20 meter standards allow the error to be only 2.0 times larger at 1%. [3]

CT phase angle error adds a second problem as power factor falls. AN-136 gives the example of a C57.13 class 0.3 CT, which allows a phase angle error of 0.25 degrees (15 minutes) at 10% of rated current; uncorrected, that causes a 0.75% error at a power factor of 0.5, which exceeds the 0.6% maximum error allowed by C12.20 accuracy class 0.5. [3]

For CT-operated meters, ask whether the stated accuracy is for the meter alone or for the meter tested with the CTs you will buy, and request CT gain and phase specifications from 1% to 100% of rated current. AN-136 notes that C12.1-2014 has accuracy classes 1.0 and 0.5, C12.20 has classes 0.5 and 0.2, and both include accuracy tests down to 1.5% or 1% of rated current. [3]

05

Ask how the current sensor behaves under 2–150 kHz disturbances

A 2026 open-access study in Sensors tested directly connected class B (1%), 5 A nominal meters with different current sensors (shunt resistor, current transformer, Rogowski coil, Hall-effect sensor and an induction meter), plus a class 0.2 digital meter fed by a merging unit with IEC 61850-9-2 sampled values, against conducted differential-mode disturbances based on IEC 61000-4-19. The authors note that the product-standard levels for directly connected meters are 3 A at 2–30 kHz and 1.5 A at 30–150 kHz, and that their screening applied 1 A over an extended 1–150 kHz range in 1% frequency steps against a reference meter with error not above 100 µWh/Wh. [4]

Under 1 A continuous-wave disturbance, the largest error deviation of the class B meters was 0.375% (induction meter) and the smallest 0.122% (Rogowski coil), while the merging-unit system reached 1.176%. Under rectangular-modulated disturbance the shunt-based meter reached up to 0.816% at 101 Hz modulation. The authors report that all conventional meters stayed within the limits for their declared 1% class at this level, and that the shunt and Rogowski-coil meters showed the lowest overall susceptibility. [4]

Editorial calculation: the 1 A test level is one third of the 3 A product-standard level at 2–30 kHz and two thirds of the 1.5 A level at 30–150 kHz, so these results are a screening, as the authors state, not a compliance result. Where a site has inverters, chargers or LED drivers on the same circuit, ask the meter supplier for its IEC 61000-4-19 test report at the product-standard levels and for the current-sensor type. [4]

06

Trace the meter construction

This is a functional reference built from the sources cited; it is not one maker's specification.

07

Fill the specification fields with values for your meter

An energy meter quotation needs values for your installation. The table explains each field, gives published values where a reviewed source states them, and says what to obtain from the supplier.

08

Send the supply, class and outputs together

Paste it into your enquiry below or into an email to a maker, then replace every bracket.

Supply: [1p2w, 3p3w, 3p4w; Un; fn]
Current: [direct Imax or CT ratio]
Accuracy: [class A/B/C; MID or other approval]
Energy: [import/export; reactive; tariffs]
Outputs: [pulse; Modbus; M-Bus; wireless]
Mounting: [DIN rail or panel; sealing]
Environment: [temperature; enclosure]

See manufacturer records Send an enquiry through CNFX