5.1.2 Potential influences/disturbances and limitations

Generally speaking, influences or disturbances are the result of the presence of or change in physical phenomena that influence the measurand, but not the measurand itself.

Spectrum of phenomena

The behavior of the phenomena can be quite diverse. The values of the climatic parameters generally do not change as fast as the electromagnetic phenomena.

Also, the nature of these physical phenomena can be quite diverse. For example some are only one-dimensional such as temperature, humidity and air pressure while others are multidimensional, such as vibration and electromagnetic phenomena. Another parameter which has to be taken into account and which differs quite a lot between phenomena is the dwell time.

This difference in behavior will influence the risk on the measurand; the risk assessment approach should therefore be fitted to this behavior.

Table 1 — Nature of phenomena that are potential sources of influence

Physical phenomenon	Range dimensions	Orientation; polarization	Frequency	Dwell time
Temperature	1			Medium-long
Humidity	1			Medium-long
Air pressure	1			Long
Vibration	4	x	x	Short-medium
Magnetic	1	x		Short
Electric	1	x		Short
Electromagnetic	3	x	x	Ultra short-short
Chemical				Long
Etc.

Actual coverage by performance requirements

Over the past 20-30 years, performance requirements for measuring instruments that are exposed to influences and disturbances on the enclosure port have been implemented in OIML Recommendations for a number of phenomena. Influences on the measurand (internal, external) have been taken into account, but the actual risk of such disturbances in most cases has not been addressed.

For phenomena having a long or medium term dwell time this risk can easily be estimated. It has more or less already been taken into account by specifying the rated operating conditions.

For short and ultra-short dwell times the risk cannot easily be estimated, but in general the appropriate requirements and test levels are copied from generic IEC standards and are based on experiences on interference in practice. Therefore, these levels should implicitly take into account the risk of interference.

For some phenomena (see below) no performance requirements have yet been specified. One could assume that disturbances as a consequence of these will be covered by the general performance requirements of the measuring device. The risk of a disturbance, however, will be unknown when the dwell time of the phenomenon is short and its existence is location dependent.

Table 2 — Actual coverage by standards referred to in OIML D 11 and in the applicable Recommendations

Infl./dist. / Port	Power supply	Measurand	Enclosure	Data transmission	Operator
Climatic/vibration
Magnetic
Electric
ESD
Interruption
Variation
Surge/burst
RF induced currents
LF induced currents
Rad. out of band
Cond. out of band

	Not applicable
	Could be applicable
	Covered in OIML D 11
	Partly covered in D 11
	No standard available
	Covered by UTC/ETSI

Low frequency phenomena
As can be seen from Table 2 for a number of relatively low frequency conducted and radiated phenomena, standards are not available and/or standardization is in progress. Caution is therefore advised and it may be necessary to develop and perform product specific tests.
Mutual interference of instruments in the same vicinity
One should be aware that the levels for EMC immunity testing as specified in generic IEC standards do not cover the risk of mutual interference between adjacent instrumentation.
It could be expected that manufacturers in such cases will notice any undesirable behavior during prototyping. But this approach would not cover the potential interference from a device from some other manufacturer, which for example would be the case when power line communication is also in use next to the smart meter data communication line.
Estimating the level of interference from radiating sources
When attempting to estimate the level of electromagnetic interference, the following parameters need to be assigned a value:
- expected environment of operation;
- level of immunity of the measuring instrument.
Both tend to be complex but considerable research has been carried out and studies provide some figures on expected maximum levels of emission of sources.
When determining the properties of the environment, all contributing sources and distances from these sources are to be taken into account.
When determining the expected level of immunity, a mathematical model could be used or immunity tests could be performed. Both could be rather complex.
Simple but adequate models and/or reproducible test setups should be created.
One approach could be to base the deduced level of interference on the maximum expected emissions and optimal coupling between the sources and the “victim”. The outcome of such an approach would probably be that some (perhaps most) of the instrumentation would not be in conformity with the requirements for non-interference.
Another approach could be to include a risk analysis, taking into account both the actual risk when a potential source is present, and the coupling factor. This factor would comprise e.g. distance, polarity and isotropy components, each of which contributes to the resulting intensity on the victim location.
Such an analysis can be performed using a mathematical model, but setting up such a model requires additional work. A Monte Carlo approach could probably be the best technique to estimate the risk.
Coverage of immunity tests
Over the past ten years an exponential increase in the use of the electromagnetic spectrum has been observed, mainly for communication purposes. This accounts for both transmission line bound and free space phenomena.
Driven by commercial incentives the telecommunication companies have optimized the use of the limited available EM bandwidth by using sophisticated and intelligent software methods.
This increase in use, however, has not kept pace with the limited bandwidth available and the techniques for exploiting the upper RF bandwidth regions; therefore, the potential risk of conflicting use of the spectrum is increasing.
For decades, standardization committees have been intensively working on preventing mutual interference by setting requirements on emission and susceptibility of (mainly) electronic devices.
Their first focus on EM interference originated from the prevention of disturbance of radio services.
Besides this prevention of evident interference risks on communication, standardization of EM interference-related qualities further arose from hazardous incidents and commercial pressures.
However, since the occurrence of an interference in many cases is stochastic in nature, protective measures and standardization of these measures have only been implemented in those cases where there is a relatively high risk of occurrence and where there are substantial consequences.
A review of the agreed measures as specified in standards over the whole EM spectrum shows that this process has lead to an incomplete coverage of EM interference protective measures, resulting in certain gaps in specific bands.
The introduction of smart metering has made these gaps in protection more manifest.
One gap which has become more prominent concerns the VLF and LF band protection. Due to the fact that in this band below 150 kHz no (efficient) radio transmissions are operated, there is no real concern and therefore no involvement of radio protection agencies in this band. The use of EM phenomena and signals in this band is merely transmission line bound and emissions to free space are beneath the level of observance (noise level) within a few meters of the transmission line.
Electricity suppliers and distributors make use of this frequency band for switching and signaling over mains power lines. Since instrumentation connected to the mains will in general frequently be exposed to such signaling and since its properties are well defined, the risk of unexpected interference will already become prominent at the design stage of such instruments, and can be reduced prior to marketing the product.
Of more concern are those mains connected products which produce disturbances in this frequency band and whose waveforms are more or less arbitrary pulses. Smart meters, when directly connected to these mains power supplies, can suffer from these kinds of signals, not only as a result of the interference on the measurand but also on the measured data when using PLT/PLC as a means of communicating this data.
No adequate measurement methods or standardization are applicable for these kinds of interferences.

Figure 2 — Overview of the electromagnetic spectrum in use for radio and infrared communication

Frequency bands in use for smart metering are:

VLF and LF (conducted) band for PLC;
UHF band (radiated) for GPRS.