This is a question that often comes up in smart metering projects, but perhaps it is not the right question to ask, because battery life is not an intrinsic characteristic of the meter.
It is the result of the interaction between technology, service configuration and actual operating conditions.
Yet, tender specifications and procurement requirements often include statements such as: “Battery life: at least 13 years”, without clearly defining the conditions required to achieve this target.
The reality is that two identical meters, equipped with the same battery and the same firmware, can have very different service lives.
Why? Because battery life depends on many factors
✅ Battery capacity
✅ Hardware efficiency
✅ Firmware optimisation
✅ Application protocol used
✅ Reading and transmission frequency
✅ Number of alarms and additional communications
✅ Radio coverage quality
✅ Installation conditions
✅ Evolution of the cellular network over time
One aspect that is often underestimated is the application protocol.
Not all protocols have the same impact on energy consumption: some are more verbose and energy-intensive than others. The more messages are exchanged, the more data are transmitted, and the longer the modem remains active, the higher the energy consumption. The choice of application protocol therefore has a direct impact on battery life.
The most important point, however, concerns NB-IoT connectivity
Unlike local radio technologies such as wireless M-Bus, an NB-IoT smart meter is part of an end-to-end solution involving at least three key players: the meter, the platform collecting the data, and the mobile network operator carrying the communications.
The overall performance of the system depends on all three.
The cellular network, in particular, is a dynamic infrastructure: it evolves over time and is continuously optimised, reconfigured and upgraded. Coverage, radio parameters and frequency usage can all change.
Many tender specifications and SLAs are written as if the radio conditions on day one will remain identical fifteen years later. Yet anyone—even someone new to smart metering—knows this is not the case: we have all experienced how mobile network performance can vary over time simply by using our smartphones.
The NB-IoT network is not a constant. It is a variable.
All these factors can influence the device’s energy consumption throughout its lifecycle.
This does not mean that the battery life of a smart meter is unpredictable or that the manufacturer is simply a passive observer. On the contrary, a significant part of the device’s service life is determined by the design choices made before the meter is ever installed in the field.
At Maddalena, we have worked on several strategies with one straightforward objective: to make the meter as resilient as possible to real-world operating conditions, optimising its efficiency across the different conditions it may encounter in the field.
Perhaps, then, the right question is not: “Will the meter last 13 years?” but rather: “Which operating conditions have been considered, and which measures have been adopted to maximise the likelihood of achieving those 13 years?”
For this reason, when evaluating an NB-IoT smart meter, battery life should always be associated with a clearly defined operating scenario: transmission frequency, application protocol, radio coverage, alarm management and installation conditions.
Only then does the figure become comparable across suppliers and genuinely useful in a technical specification or purchasing decision.
Because a declared battery life without reference conditions is just a number.
A declared battery life together with its operating conditions is a technical requirement that can be measured and verified.
And ultimately, battery life is not guaranteed simply by using a larger battery.
It is engineered by designing a system capable of making better use of every available joule.
