How to Size a Telecom Backup Battery: Site Autonomy
How to Size a Telecom Backup Battery: Site Autonomy
Telecom sites sit in places where grid power is unreliable or absent, and the battery bank is what keeps a tower or cabinet live through an outage. “Site autonomy” is the number of hours the bank can carry the load with no grid input, and sizing it correctly is a balance between uptime, cost, footprint and maintenance. Undersize and the site goes dark mid-storm; oversize and you have paid for capacity that never cycles. A lithium battery manufacturer experienced in telecom will start from your load profile and target autonomy, then work backward to capacity, not the other way around.
Define autonomy: hours, not just capacity
Autonomy is expressed in hours of backup at the site’s critical load. A rural macro tower might need 8 to 12 hours to ride a storm; a cabinet with generator backup might need only 2 to 4 hours to bridge the start gap. Start by deciding the outage you must survive—often driven by local grid history or a service-level agreement—then multiply by the load. Treat autonomy as a design input, not a number you discover after buying cells. A battery application solutions plan that skips this step ends up either dark or wasteful.
Measure the real load, not the nameplate
The site load is the sum of radio units, baseband, rectifiers’ overhead, cooling and any auxiliary gear, measured in watts or amps at the bank voltage. Nameplates overstate; measure at the DC bus over a day to capture diurnal variation and the extra draw of cooling on hot afternoons. A site quoted at 500 W might actually pull 700 W at 35 C once the fan runs full, and that 40 percent gap is exactly what strands a marginally sized bank. Log the load for a week if you can; the average and the peak both matter for different reasons.
The autonomy capacity formula
| Step | What you compute | Example (700 W, 48 V, 10 h) |
|---|---|---|
| Load current | Watts divided by bank voltage | 700 / 48 is about 14.6 A |
| Ah at autonomy | Current multiplied by hours | 14.6 x 10 = 146 Ah |
| Depth-of-discharge allowance | divide by usable DoD (0.8) | 146 / 0.8 = 183 Ah |
| Efficiency and margin | multiply by 1.1 to 1.2 | 183 x 1.15 is about 210 Ah |
The table turns autonomy into a concrete capacity. For a 48 V site needing 10 hours at 700 W, you land near 210 Ah of nameplate to keep the bank above its 80 percent depth-of-discharge floor through the outage. Note this assumes lithium; a lead-acid bank would need far more because its usable DoD is only about 50 percent, which is why lithium has taken over telecom backup despite higher upfront cost.
Lithium versus lead-acid for telecom
Lead-acid has ruled telecom for decades because it is cheap per amp-hour, but it suffers on depth-of-discharge, cycle life and maintenance. Lithium LFP delivers 80 to 90 percent usable DoD versus lead-acid’s 50 percent, doubles or triples cycle life, and needs no watering or equalization. The result is a smaller, lighter bank that fits a cabinet and survives more outages. A sodium-ion battery is emerging as an even cheaper stationary option for cold or cost-first sites, though its field record in telecom is younger. The decision is increasingly lithium by default, with sodium a watch item as its certs mature.
Temperature and enclosure
Battery capacity and life both fall with heat, and telecom cabinets in direct sun can bake. Size the bank with the worst-case cabinet temperature in mind, and if cooling is intermittent, add capacity margin or active ventilation. Lithium tolerates a wider range than lead-acid but still ages faster above 45 C, so the enclosure design is part of the sizing, not an afterthought. A home energy storage system in a climate-controlled room has it easy; a roadside cabinet does not, and the bank must carry that reality.
BMS, monitoring and remote alarm
A telecom bank you cannot see is a bank that fails silently. Require a BMS that reports per-cell voltage, temperature and state-of-charge over the site’s management bus, and an alarm that pages you on low autonomy before the load drops. Remote telemetry lets a control center reroute or dispatch before a site goes dark, turning a failure into a managed event. For fleets of hundreds of sites, this visibility is the difference between a handful of embarrassing outages and a regional blackout during a storm. The BMS is not optional hygiene here; it is the operational backbone.
Generator bridging and hybrid sites
Many sites pair a short battery autonomy with a generator or a renewable source. If a genset starts within minutes, you may only need 1 to 2 hours of battery to bridge the gap, slashing capacity and cost. Solar-hybrid sites size the bank to cover the night plus a cloudy-day margin, with the panels doing the daytime work. The autonomy target drops sharply once another source is in the mix, which is why the first question is always “what else keeps this site up?” before you commit to a big bank. A semi-solid-state battery is rarely the right call for stationary telecom because its density premium buys nothing in a floor-mounted cabinet, so plain LFP or sodium wins on value.
Redundancy and single points of failure
For critical sites, a single bank is a single point of failure. Consider N+1 battery strings, dual rectifiers, and diverse feeders so no one fault darkens the site. Redundancy costs more but the cost of a dark macro tower—dropped calls across a region, emergency services blind—is far higher. Size each string to carry the full load for the autonomy target so a failed string does not immediately strand the site. This is standard practice for tier-one operators and should be in your spec from the start, not bolted on after an incident.
Commissioning and capacity verification
Do not trust the nameplate at commissioning. Perform a discharge test that loads the bank to the real site current and confirms it holds voltage and capacity through the designed autonomy, ideally with the BMS logging. A bank that passes on paper but sags under the actual load reveals manufacturing or integration faults before the storm does. Repeat the test annually; lithium holds capacity well, but a faulty cell or loose connection shows up as early autonomy loss you can catch and fix cheaply.
Procurement checklist
- Set the autonomy target from grid history or SLA, not a guess.
- Measure the real load including cooling at worst-case temperature.
- Compute capacity with DoD, efficiency and a design margin.
- Choose lithium, or pilot sodium, over lead-acid for DoD and life.
- Require BMS telemetry and remote low-autonomy alarm.
- Commission with a real discharge test, not a paper check.
Cell balancing and string health
In a telecom string of many cells in series, imbalance slowly caps usable capacity and flags false low-autonomy alarms. A BMS with active or adequate passive balancing keeps the string even, and periodic manual top-balancing during service extends life. For large sites with multiple strings, monitor per-string current so one weak string is not silently carrying more than its share. The cheapest capacity you own is the capacity you stop losing to imbalance, and in a remote site where a truck roll is expensive, that discipline pays for itself quickly.
Site classification and tiering
Not every site deserves the same autonomy. Tier-one macro sites carrying public-safety traffic warrant 12 or more hours and full redundancy; a microcell in a powered building may need only minutes. Classify sites by traffic importance and grid quality, then assign autonomy and redundancy accordingly so your budget goes to the sites whose failure hurts most. This tiered approach prevents over-spending on a cabinet nobody notices while under-protecting a node that carries a region, which is the classic way backup budgets get wasted.
Replacement and end-of-life planning
Plan the battery’s swap before you need it. Track state-of-health through the BMS and set a threshold—say 80 percent remaining capacity—at which you schedule replacement during a planned visit rather than after a failure. Lithium’s long life means a site may go years between swaps, so the replacement budget and logistics are easy to forget; document the date and train local staff to read the alarm. A bank that fails at 2 a.m. during a storm is far costlier than one swapped on a sunny afternoon on schedule, and the planning is half the reliability battle.
Frequently asked questions
How many hours of autonomy do I need? It depends on grid history and SLA; 8 to 12 h for gridless sites, 2 to 4 h where a generator bridges.
Why not just oversize the bank? Oversizing wastes money and footprint; size to the target plus margin, not blindly large.
Is lithium worth it over lead-acid? Yes for telecom: 80 percent plus usable DoD, longer life, no maintenance, smaller cabinet.
Should I add a generator? If a genset starts in minutes, you can cut battery autonomy to 1 to 2 h and save capacity.
Written by Karl at China Battery Technology. Give us your load and target autonomy and we will size the bank—request a quote for a telecom-ready system.
Related Reading
- battery application solutions hub
- How to Choose a Marine Battery: Vibration and Humidity
- How to Size a UPS Battery: Bridge Time Math
