Menu
Application GuideSelection Guides16 min read

When Should You Replace a Soil Moisture Sensor?

There is no universal replacement date. This guide separates routine inspection, recalibration, repair and replacement using four lifecycle clocks, then covers the review cadence, warranty limits and a spares plan built around true interchangeability.

Updated October 4, 2026How Equipvia researches
Digital multimeter used to check a soil sensor signal path before replacement
Replace on evidence, not on the calendar: prove the signal path healthy first, then separate recalibration from replacement and hold tested spares. Digital multimeter photograph via Wikimedia Commons (CC BY-SA 2.5). View current BGT listing.

Disclosure: We may earn a commission if you purchase through links on this page, at no additional cost to you.

There is no universal replacement date. A soil-moisture sensor should not be replaced simply because a calendar says it is two or three years old. Service life depends on sensing technology, soil chemistry, burial conditions, cable damage, battery and communications design, required accuracy, and whether the supplier still supports the exact model. A sensor can remain physically intact while its output becomes unsuitable for a control decision. Another can be working correctly but become a procurement risk when its protocol, spare parts or documentation is no longer available.

The practical buyer task is to separate four decisions: routine inspection, recalibration or validation, repair, and replacement. Campbell Scientific notes that soil-moisture probes typically do not need routine recalibration after installation, while contaminated sites may justify more frequent checks [E01]. Other guidance documents describe sensor-specific replacement intervals, such as three to four years for some gypsum-based sensors, but that interval cannot be transferred to an FDR, capacitance or SDI-12 probe [E02].

Use four clocks instead of one service life number

A lifecycle plan becomes more useful when it tracks different failure clocks. The clocks can reach a decision at different times, so a sensor should be replaced when the first relevant limit is reached—not when a generic age threshold is reached.

Lifecycle clock What changes Evidence that should trigger action
Functional clock The device no longer powers, communicates or returns a stable record. Repeated missing packets, failed self-checks, damaged cable, water intrusion, unstable supply or a failed output test.
Measurement clock The device still reports but no longer supports the required decision quality. Persistent offset against a reference, unexplained drift, implausible event response or a failed site validation.
Environmental clock The installation has changed the conditions for which the sensor was selected. Corrosion, damaged probe surface, compacted or washed-out soil contact, fertigation exposure, repeated mechanical disturbance or connector intrusion.
Commercial clock The buyer can no longer obtain a defensible, compatible or supportable replacement. Discontinued model, unavailable calibration file, changed register map, no replacement probe, long lead time or unclear warranty route.

The commercial clock is easy to miss. A low-cost sensor may still produce numbers, but if the supplier cannot provide the same output scaling or model revision, a replacement can create a break in the time series. Record the model, serial number, firmware, configuration, interface and installation date so a future replacement is a controlled change rather than a silent substitution.

Inspect first, then decide what kind of intervention is justified

A failed reading is not proof that the probe itself is bad. A loose terminal, a damaged cable, a wrong register map, an air gap or a gateway problem can create the same symptom. Conversely, cleaning a connector will not repair a sensing element that has a repeatable bias. Use a staged decision so that replacement is not used to hide an installation or configuration problem.

Observation Most defensible first action When replacement becomes justified
No data from one node Check power, cable continuity, connector, logger channel, address and communications before removing the probe. The signal path is proven healthy and the probe fails a documented output or bench check.
Stable but implausible values Compare a reference measurement at the same depth and inspect soil contact, scaling and configuration. The error repeats after installation and configuration checks, and recalibration or a site curve cannot restore the required decision quality.
Slow drift across one sensor Review temperature, salinity, soil movement, firmware and the reference record; preserve the raw data. The drift is sensor-specific and remains outside the acceptance limit after a controlled validation.
Physical damage or water intrusion Isolate the channel, photograph the condition and inspect the complete cable and connector chain. The enclosure, sensing surface or sealed transition has lost its environmental boundary or cannot be repaired with documented parts.
Model discontinued or changed Obtain the last datasheet, register map and replacement notice; test a proposed successor in parallel. The successor cannot reproduce the required output, documentation or data continuity within the project's acceptance criteria.

The first action should be recorded even when the final decision is replacement. This preserves the failure history and prevents the same wiring, placement or configuration mistake from damaging the next unit. NRCS maintenance guidance makes the same operational point: buried sensors are difficult to inspect without disturbing the site, so data reasonableness and collocated comparisons are important, and malfunctioning sensors may need to be removed and replaced rather than repaired in the field [E03].

Set the review cadence from the risk, not from habit

A maintenance schedule should reflect the consequence of a bad reading and the cost of reaching the station. A research plot with a recoverable data logger can use a different cadence from a remote irrigation block whose readings open valves automatically.

Review point Checks to perform Record for the lifecycle file
Before each season or deployment Confirm model, serial number, firmware, configuration, battery, connector condition and reference check. Baseline output, installation photograph, configuration export, date and responsible person.
Routine remote review Look for missing data, flat lines, impossible values and stale timestamps; compare against rainfall and irrigation. Flagged periods, reviewer, decision and follow-up action.
After a wetting event Confirm the response arrives at the expected depth and time at this station. Event timestamp, applied volume or rainfall and the observed response.
Periodic field visit Inspect soil contact, cable, connector, conduit and enclosure; compare a reference measurement. Photographs, reference readings, soil condition and any correction applied.
Before model or supplier change Test the proposed successor in parallel against the incumbent. Parallel dataset, variance statement and the date the new series becomes authoritative.

A useful lifecycle rule names a decision point when the measurement matters, rather than promising a fixed number of years.

Warranty is not service life

A warranty describes the supplier's remedy under stated terms. It does not promise that the measurement remains within a buyer's accuracy limit for the same period, and it does not prove that a sensor can be repaired in the field. Ask whether the warranty covers the probe, cable, connector, logger, battery and shipping separately. Also ask whether replacement units retain the same firmware, output scaling and calibration file.

A replacement policy is more valuable when it names the exact model and gives the buyer a way to maintain data continuity. Agriculture Victoria advises buyers to consider the maintenance required and whether a replacement product is available, not only the initial specification [E05]. For a remote network, the lead time for an identical spare can matter more than a small difference in purchase price.

Plan spares around interchangeability

Keep at least one tested spare for a critical station when the lead time, access cost or crop consequence is high. The spare should not be treated as "compatible" because its connector looks the same. Verify the measurement principle, probe geometry, output scale, protocol, address settings, supply voltage, cable pinout, firmware and platform mapping.

  • Maintain a golden configuration: current datasheet, register map, firmware version, calibration file, wiring diagram and installation photograph.
  • Bench-test each spare before storage and record its serial number, baseline output and date of test.
  • Use a parallel deployment when changing model, supplier or sensing principle; do not splice the new series into the old one without a change point.
  • Set a maximum acceptable time for a failed station to remain out of service. That limit should drive spare quantity and supplier lead-time requirements.
  • Ask whether the sensing probe can be replaced separately from the logger or gateway. A modular replacement can reduce downtime, but only if identity and calibration remain controlled.
  • Retain the failed unit until the root cause is documented. Destroying it immediately removes evidence needed for a warranty claim and fleet-level learning.

Two sourcing candidates for different replacement plans

These offers are sourcing candidates for lifecycle planning, not certified recommendations. The marketplace fields below are seller-stated unless explicitly identified as a landing-page or supplier-host observation. The two listings share a BGT supplier, so the supplier inquiry route is intentionally shared; each product link points to a different listing. The second listing uses "Teros12 5TE" wording, but the listing does not independently establish original METER Group identity or equivalent manufacturer provenance.

Candidate — BGT Smart CE 3-in-1 FDR soil sensor

Candidate — BGT listing marketed as Teros12 5TE

Write the replacement rule before the first failure

A useful replacement rule is short enough for a field technician to apply and precise enough for a data manager to audit. For example: remove a sensor from automatic control when it has a confirmed communication failure, visible environmental damage, or a validated measurement error outside the project limit. Keep the raw record and flag the affected period. Recalibrate only when the sensor and installation remain physically sound and the manufacturer's method supports it. Replace when the fault is in the sensing element, environmental seal, undocumented variant or unsupported model.

The rule should also say who can return a sensor to service. A technician may restore a connector or replace a battery; a change of probe model, firmware, register map or calibration equation should require technical review. This prevents a well-intentioned spare swap from creating a hidden discontinuity in irrigation or research data.

Bottom line

Replace a soil-moisture sensor when the evidence shows that the measurement or the supportable installation has crossed its acceptance boundary—not merely because the calendar changed. Inspect the signal path first, separate recalibration from replacement, document the model and configuration, and hold tested spares where downtime matters. The two BGT listings are useful sourcing leads for different replacement plans, but their warranties and marketplace specifications must be converted into an exact-variant RFQ and a parallel field pilot before they are treated as interchangeable fleet components.

Evidence and source notes

Share this article

Related research

View all guides
Cooling & Condensation·20 min read

How to Size a Chiller for a Rotary Evaporator

Size a rotary evaporator chiller by required condenser temperature and cooling capacity at that temperature—not by minimum temperature or flask size alone.

Updated September 19, 2026