The End of 2G and 3G: How Can You Prevent Fuel Sensors from Stopping Working?

Reading time: 10 min. · Updated on July 15, 2026 · Prepared with the expertise of François Clément, Sales Director at FOUR DATA

• When will 2G and 3G be phased out in France?
• And what about the rest of Europe? Seven countries, seven calendars
• Why Your Tank Sensors Are on the Front Lines
• Taking Inventory of Your M2M Devices: A 5-Step Method
• Prioritize replacements based on business risk
• LTE-M, NB-IoT, or LoRaWAN: Which connectivity technology will follow 2G?
• How FOUR DATA Ensures the Continuity of Oversight
• What Will Change After the Migration Is Complete
• FAQ
• Is your fleet ready for September 2026?

The end of 2G is no longer a distant prospect: Orange will shut down its 2G network in nine departments in the southwest between March and June 2026, and between September 22 and October 20, 2026, the shutdown will be extended to all of metropolitan France. SFR and Bouygues Telecom will follow suit by December 2026.

For an energy supplier or industrial operator, any meter, modem, or level sensor still equipped with a 2G module will stop working on the date the network is shut down in their region. This guide will help you take inventory of your equipment, prioritize replacements, and choose a sustainable connectivity path.

2G will be shut downin mainland France by the end of 2026: Orange will carry out the shutdown between September 22 and October 20, 2026, while SFR and Bouygues Telecom will shut down their 2G networks by December 2026. Next will be 3G: by the end of 2028 for Orange and SFR, and by the end of 2029 for Bouygues Telecom.

Two details make a big difference for a distributed M2M network. First, the shutdown occurs zone by zone, not all at once across the entire country: a sensor in Bayonne stopped transmitting in March 2026, while its counterpart in Lille will remain operational until October.

Next, the planning is in the hands of the network operators. Arcep publishes and updates the information on its fact sheet regarding the shutdown of the 2G and 3G networks; this information must be reviewed before a decision is made.

Sources: ARCEP and national providers, overview by FOUR DATA, updated in June 2026. Timelines are subject to change.

There is no common European deadline: each operator sets its own timeline, and the difference amounts to as much as five years between Sweden—which will phase out coal-fired power as early as December 2025—and Belgium, which has given itself until 2030.

Three situations deserve special attention.First, Germany: the shutdown for the general public will take place in 2028, but Vodafone will maintain a 2G service reservedfor critical IoT applications (elevators, eCall, alarm systems) until the end of 2030. This extension applies only to specific contracts; a standard fuel tank sensor does not automatically qualify.

Next is Belgium, where the order is reversed: 3G has been shut down there since mid-2025, while 2G will remain in use until 2030 at the latest. So a 3G modem won’t work there anymore, while its 2G counterpart still does. Finally, Denmark: no provider has announced a date, which offers no guarantee of its continued existence. If new 2G equipment were to be deployed there today, it would amount to installing a technology that has already been written off elsewhere in Europe.

For a distributor operating in France, the Benelux, and the DACH region, this disparity means that a retroactive schedule must be drawn up for each country, rather than a single overall plan. FOUR DATA is deploying its LTE-M, NB-IoT, and LoRaWAN sensors in these various markets and aligning the replacement phases with each operator’s local schedule: first France and Sweden, then the Netherlands and Norway, and in a second phase, Germany and Belgium.

A 2G smartphone stands out: you can’t make calls with it anymore, the owner complains. A level sensor installed eight years ago on an LPG tank in a rural area isn’t sending a signal. It stops transmitting the measurement data, the platform displays the last known value, and no one notices until the next inspection—or, worse yet, until a leak occurs at the customer’s site.

Arcep’s figures confirm the extent of the lag. As of the end of March 2026, Arcep’s quarterly report still listed 2 million SIM cards in devices that support only 2G, and 4.5 million when 2G/3G devices are included. A significant portion of this total consists of M2M devices: emergency alarms, meters, trackers, and, more specifically, fuel level sensors and telemetry modems installed in the early 2010s.

These devices fall outside the scope of traditional IT inventories for one simple reason: they don’t belong to anyone. The tank is located at the end customer’s site, the sensor was installed by a subcontractor, the SIM card is managed by an external M2M operator, and the subscription has been automatically renewed for years. The result: on the day of the outage, the distributor loses track of dozens or hundreds of inventory locations without any prior warning.

At FOUR DATA, we have assisted several distributors with this inventory process. Time and again, the same pattern emerges: the actual number of installations turns out to be higher than initially estimated, because the oldest installations are also the least well-documented.

1. Compile a list of active SIM cards for each operator or M2M connectivity provider, specifying the access technology (2G, 3G, 4G, LTE-M, NB-IoT).
2. Reconcile with the list of field equipment: meters, modems, control units, GPS trackers, including serial numbers, installation dates, and locations.
3. Determine which devices support only 2G and which support both 3G and 2G: according to the providers’ schedule, these will be phased out.
4. Assess each location: stock criticality, distance, accessibility, whether or not it is a strategic customer, and the presence of an ATEX zone.
5. Draw up the replacement schedule, starting with the zones whose expiration dates are closest.

In Step 4, an Excel file is converted into an action plan. Based on the company-specific assessment, we first address the areas that will cost a lot of money if no action is taken: the customer’s tank with a contract for automatic restocking, the remote location a 90-minute drive away, and the classified warehouse.

Would you like to outsource this assessment? The teams at FOUR DATA will compare your list of locations with the outage schedules and network coverage tests.

Not all tanks are the same. In most parks we inspected, a three-tier classification is sufficient.

Level 1 must be addressed before the zone is closed off. Locations where sensor failure poses an immediate risk: a gas leak at a business customer’s site, flooding at an ICPE site, or a GNR tank on a farm during harvest. At these locations, replacement is scheduled as a priority intervention, not as a routine maintenance task.

Level 2: Complete within 6 months. Locations that are monitored remotely to optimize delivery routes, without posing an immediate risk of stock shortages. Data loss in this context leads to a decline in logistics performance (the need to call customers back, incompletely filled delivery routes), without causing a safety crisis.

Level 3, to be evaluated. Equipment that has reached the end of its commercial life, canceled customer accounts, and removed tanks. The phase-out of 2G provides an opportunity to clear out the equipment inventory and stop paying SIM subscription fees for inactive connections.

This sorting process has a useful side effect: it reveals the true costs of the legacy equipment fleet. Several of our customers have discovered in the process that they were being billed for dozens of M2M lines for equipment that hasn’t been in use for years.

There are three technologies used to replace 2G modules in tank sensors. The right choice depends on the location, not on a general preference.

A theoretical calculation is no substitute for an on-site measurement. LTE-M or NB-IoT coverage varies from one technical area to another, even within the same municipality.

FOUR DATA designs and manufactures industrial IoT sensors that communicate natively via LTE-M, NB-IoT, LoRaWAN, or Sigfox, and integrates them with its monitoring platforms, Desk and Sens. The NOVA product line covers the key applications of the migration: NOVA PRESSION for tanks containing diesel, GNR, fuel oil, and AdBlue; NOVA ROCHESTER for tanks equipped with Rochester level sensors; and NOVA PULSE for gas meters.

There are approximately 80,000 FOUR DATA sensors in use in Europe, at distributors such as Antargaz Energies, Vitogaz, Dyneff, and Picoty. This existing installed base is of direct relevance to a migration project: the various scenarios (different tank thread types, areas without coverage, ATEX locations, and mechanical level gauges that must be retained) have already been identified and resolved, often using in-house-manufactured adapters.

The migration is not limited to replacing a housing. Replacing a 2G sensor is the perfect opportunity to review the entire chain: a reporting frequency tailored to usage (data is transmitted at a configurable frequency, ranging from several times a day to once a week, depending on the criticality), alert thresholds, integration of the levelsinto the ERP system or route planning software, and a white-label customer portal via Sens for distributors who want to turn this technical challenge into a visible service offering.

The first advantage is defensive: data continues to come in even after a failure. The second advantage is offensive, and this is often the argument that justifies the budget. The latest sensors provide more reliable and frequent measurement data than the previous generation’s 2G modems. The teams are shifting from reactive management (delivering when the customer calls) to proactive management (delivering when the autonomy forecast recommends it).

Let’s be clear about what we’re promising: the migration does not eliminate the need for interruptions or interventions. It reduces uncertainty, makes route planning more reliable, and gives customer service a head start on incoming calls. For a fleet of several hundred tankers, this difference translates into more fully loaded trips and the prevention of emergencies.