Skip to main content

Compare

Energy meter accuracy class: 0.5S, 1 or 2

The number is the error band. The letter is the part most buyers miss — and the part that decides whether the meter is still accurate at the loads your circuits actually run at.

The short answer

If the number goes on an invoice, in an audit report, or into a published metric — specify Class 0.5S. Billing, sub-billing, measurement and verification, PUE: all of these get questioned, and a lower class gives the person questioning them a valid argument.

Class 1 is reasonable for general monitoring and trending, where you want to know that something is running when it should not be.

Class 2 is indication. Treat its output as a direction, not a measurement.

Side by side

What each class is fit for, rather than what its datasheet says.

ConsiderationClass 0.5SClass 1Class 2
Governing standardIEC 62053-22IEC 62053-21IEC 62053-21
Nominal error band±0.5%±1%±2%
Accuracy held at low loadYes — the S is for specialReducedReduced
Suitable for tenant billingYesMarginalNo
Suitable for energy auditsYesIndicative onlyNo
Suitable for PUE reportingYesQuestionableNo
Suitable for trend monitoringYesYesYes
Survives a billing disputeYesSometimesNo
Typical useRevenue and sub-billing, M&VGeneral monitoringIndication only

The S is doing most of the work

Comparing 0.5 to 1 as if the only difference were half a percent misses the point. The S suffix in Class 0.5S denotes a special class specified to hold its accuracy down to a much lower fraction of rated current.

Above about 5 percent of rated current, Class 1 and Class 0.5S both hold their full stated accuracy — the S changes nothing there. It matters below that. Take a feeder on 1000/5 A CTs that draws 30 A at night: it is running at 3 percent of rated current. That is inside the 1–5 percent band where a Class 0.5S meter is still specified (with a wider limit than at full load) and where an ordinary class is either outside its specified range or carries a larger permitted error. Circuits sized for peak spend real hours in that band — nights, weekends, standby.

So the practical question is not “how accurate is this meter at full load” but “how accurate is it at the load my circuit actually sees”. That is what the letter answers.

Specify by what the number is for

Tenant or departmental billing → 0.5S

Anything invoiced will eventually be challenged. A recognised accuracy class is what turns your reading into evidence.

Energy audits and M&V → 0.5S

You are measuring a difference between a baseline and a post-retrofit case. If the difference is smaller than your measurement error, the saving is unprovable.

PUE and efficiency ratios → 0.5S on both sides

Dividing two measurements compounds their errors. Metering one side accurately and the other loosely gives you a precise-looking number that is not defensible.

General monitoring and alerting → Class 1 is fine

If the purpose is spotting anomalies and idle load, Class 1 does the job at lower cost.

Match the CTs to the meter

A transformer-operated meter is only as good as its CTs. Pair a Class 0.5S meter with Class 0.5S CTs sized so the normal load sits well inside their range.

The Titan series of energy meters is designed to Class 0.5S for active energy under IEC 62053-22, and is used with external Class 0.5S CTs.

Frequently asked questions

It means the meter measures active energy to within plus or minus 0.5 percent under the reference conditions defined in IEC 62053-22. The S suffix stands for special, and it is the important part: it denotes a class whose accuracy is specified down to 1 percent of rated current, lower than an ordinary class. So a Class 0.5S meter is not simply twice as accurate as a Class 1 meter — its accuracy is also specified across a wider low-current range.
The error band at nominal load is the same. The difference is at low load. Class 0.5S meters are specified down to 1 percent of rated current, which matters on circuits that spend long periods lightly loaded — nights, weekends, standby — at a few percent of their CT rating. If you are buying for billing or measurement and verification on circuits like that, the S is the part you are paying for.
Class 0.5S. Anything you invoice against needs to survive being questioned, and a Class 1 or Class 2 meter gives a tenant a legitimate argument. The cost difference between classes is small relative to a single billing dispute, let alone a year of them.
Less than people assume, if all you want is trends. A Class 2 meter will show you that a compressor runs at night when it should not. But the moment the data is used to allocate cost, verify a saving, or report a ratio like PUE, class becomes the thing that decides whether anyone believes the number.
Because PUE is a ratio of two large numbers producing a small one, typically between 1.3 and 1.8. Measurement error on either side moves the result materially, and in the wrong direction it can turn a genuine efficiency improvement into something indistinguishable from noise. If you intend to publish PUE, measure both sides at Class 0.5S.
No. The class is a specification under the reference conditions of the standard. In the field, accuracy also depends on operating conditions — temperature, harmonics, load variation — and on the CTs and wiring the meter is connected to. If a reading underpins a bill or a contract, agree a periodic verification against a reference meter with the other party.
It is better, but not always worth specifying. For indication and trending, Class 1 is usually fine and cheaper. The rule that holds up: if a number is going to appear on an invoice, in an audit report, or in a published metric, specify Class 0.5S. Otherwise Class 1 is a reasonable place to save money.

Not sure which class your application needs?

Tell us what the readings will be used for — billing, audit, reporting or monitoring — and we'll tell you what to specify.