Summary: For a wireless sensor, sleep current dominates battery life — a sensor spending 99% of its time in deep sleep uses almost all its energy in a few seconds of activity per cycle. The three levers that determine battery life are: reporting interval (the biggest lever), radio transmission time, and sleep current. Calculate battery life by profiling each operating state (sleep, sense, compute, transmit), computing the average current per cycle, and dividing usable battery capacity by that average. Always derate battery capacity (85% or less of nominal) and account for self-discharge — a nominal 19,000 mAh lithium thionyl chloride cell may deliver 12+ years in a quiet deployment but only 3–4 years in a busy one.

1. Why Most Battery IoT Deployments Fail Early

The most common cause of premature battery failure in IoT deployments is not a bad battery — it is a bad duty cycle. Devices designed for mains power stay awake most of the time: maintaining a wireless connection, polling sensors frequently, keeping a protocol connection alive. On battery, that behavior drains a cell in days.

A well-designed wireless sensor node wakes briefly, reads its sensors, transmits, and returns to deep sleep. The ratio of active time to total cycle time — the duty cycle — is the single most important metric for battery-powered nodes. A typical node is active only 1–5% of the time; the rest is ultra-low-power sleep.

2. The Duty-Cycle Power Budget

Battery life is accounting. For each operating state, multiply current by duration to get the charge consumed per cycle, sum the states, then divide usable battery capacity by the average current.

A representative cycle for a wireless sensor node:

The key insight: wireless transmission consumes 10–100 mA compared to 10–30 µA in deep sleep — a difference of roughly 10,000×. Radio-on time is the dominant energy cost, which is why minimizing transmission frequency and duration is the most effective optimization.

Worked Example

Consider a sensor node transmitting a 2-second radio burst (170 mA) every 15 minutes, with a 3-second sensor read (30 mA) and deep sleep (10 µA) for the rest of the cycle:

• Sensor read: 30 mA × 3 s = 0.025 mAh per cycle
• Radio transmit: 170 mA × 2 s = 0.094 mAh per cycle
• Sleep: 0.01 mA × 895 s = 0.0025 mAh per cycle
• Total per cycle: ~0.12 mAh; 96 cycles/day ≈ 11.7 mAh/day
• A 2000 mAh battery → 2000 / 11.7 ≈ 171 days

Extending the reporting interval from 15 minutes to 2 hours (with buffered transmission) cuts transmissions from 96/day to 12/day — extending battery life by roughly 3–4×, because radio time is the dominant energy cost.

3. The Three Levers That Determine Battery Life

Lever 1: Reporting Interval (the Biggest Lever)

Every transmission costs fixed energy in radio-on time. Halving the reporting frequency roughly doubles battery life; extending it tenfold extends life severalfold. The trade-off is data temporal resolution. For structural monitoring, where deformation develops slowly, hourly or even daily reporting is often sufficient.

Buffering is the key technique: collect multiple readings and transmit them in one batch. Eight readings sent in one transmission cost far less than eight separate transmissions, because each transmission carries fixed overhead (radio wake, connection, and transmit).

Lever 2: Radio Transmission Time

Transmission power and time both matter. LoRaWAN Class A is among the most power-efficient long-range options — the device opens receive windows only immediately after an uplink and sleeps between transmissions. Cellular IoT (NB-IoT, LTE-M) supports PSM (Power Saving Mode) and eDRX to reduce average current, but peak transmit current reaches 200–500 mA, requiring a battery with low ESR.

Practical steps: use the lowest transmit power that maintains link margin, keep packets short, and minimize retries by ensuring good link quality at the mounting position.

Lever 3: Sleep Current

Sleep current dominates total energy when active time is already small. If sleep current is 100 µA instead of 20 µA, the node burns roughly 4× more power before a single transmission happens. When evaluating a sensor, the datasheet’s sleep current is as important as its transmit current.

4. Battery Chemistry and Capacity Derating

Battery chemistry matters for long outdoor deployments:

• Lithium thionyl chloride (Li-SOCl₂): high energy density, ultra-low self-discharge (~2–3%/year), wide temperature range (−40°C to +85°C). The standard choice for multi-year remote sensor deployments.
• Li-ion: higher current capability but higher self-discharge and temperature sensitivity; better suited where recharging is possible.

Critical rule: never calculate battery life using nominal capacity. Derate to usable capacity — typically 80–85% of nominal — to account for cutoff voltage, temperature effects, battery aging, and reserve margin. Then apply self-discharge to get a realistic field estimate.

5. Real-World Example: The Same Battery, Two Different Lives

Consider a 19,000 mAh lithium thionyl chloride D-cell (usable ~14,800 mAh after 85% derating) in two deployment profiles:

The pattern is clear: upload frequency is the biggest lever. More uploads mean more radio time and faster battery drain — the same hardware can deliver 12+ years or 3–4 years depending on configuration.

6. Engineer’s Checklist for Battery-Powered Deployments

• Profile every state: measure sleep, sense, compute, and transmit currents — don’t rely on datasheet averages alone
• Verify sleep current: it dominates total energy once active time is small
• Set the reporting interval to the slowest that still meets your monitoring needs
• Batch readings and transmit in fewer, larger packets
• Derate battery capacity (80–85% usable) before promising field life
• Account for self-discharge and temperature effects on the chosen chemistry
• Verify link margin at the final mounting position to avoid retries and re-transmissions
• Check real-world intervals, not headline figures: a 5-year claim at 1 report/hour is not the same as 5 years at 1 report/day

7. Frequently Asked Questions

1. How do I estimate battery life for my sensor?

Profile the current of each operating state (sleep, sense, compute, transmit), compute the charge per cycle (current × duration), sum for the cycle, derive average current, then divide usable battery capacity by that average. Derate capacity to ~80–85% of nominal and account for self-discharge.

2. What is the single most effective way to extend battery life?

Extend the reporting interval. Every transmission carries fixed radio overhead, so reducing transmission frequency has the largest impact on battery life. Combined with buffering multiple readings into fewer transmissions, this can multiply battery life severalfold.

3. Why does a sensor’s sleep current matter so much?

A sensor spends most of its life in deep sleep. If sleep current is 100 µA instead of 20 µA, the node burns roughly 4× more power before any transmission happens. Once active time is already small, sleep current becomes the dominant term in the power budget.

4. Why is my battery life shorter than the datasheet suggests?

Datasheet battery life is usually calculated at nominal capacity with idealized conditions. Real deployments lose capacity to self-discharge, temperature effects, cutoff voltage, and battery aging. Also verify that the claimed life corresponds to your actual reporting interval — a claim at 1 report/hour does not apply at 1 report/minute.

5. Which battery chemistry is best for multi-year remote sensor deployments?

Lithium thionyl chloride (Li-SOCl₂) is the standard choice: high energy density, ultra-low self-discharge (2–3%/year), and a wide operating range (−40°C to +85°C). Li-ion suits shorter cycles or where recharging is available.