A 100 Ah label does not tell a telecom operator how long a battery will carry the site. It also says nothing about usable energy at the end of life, peak-current capability, cold-weather charging, parallel-rack behaviour or what happens when BMS communication fails.
The short answer: select lithium batteries for telecom sites by service duty first, then verify usable energy, power, voltage compatibility, environmental limits, protection, monitoring and lifecycle evidence. Chemistry and nameplate capacity are inputs, not a complete specification.
This guide draws on the 2025 ITU White Paper on Lithium Batteries for Telecom Sites, together with established safety and transport references. The white paper makes a useful central argument: battery quality has to survive the full chain from material selection and manufacturing to storage, transport, installation and operation. It is an industry reference, not a product certificate or a substitute for local requirements. Where this article adds engineering judgement or a worked example, that distinction is stated.
Start with the battery’s actual job
A battery intended for rare backup events is not automatically suitable for daily solar cycling or diesel-runtime reduction. Before requesting a quotation, define the load profile, required autonomy, expected charge and discharge frequency, available recharge power, reserve state of charge and the conditions under which the generator or grid should start.
| Duty | Operating pattern | Design emphasis |
|---|---|---|
| Backup-only | Long standby periods with occasional outage discharge | Calendar life, reserve confidence, self-discharge, recovery and periodic testing |
| Solar cycling | Regular overnight discharge and daytime recharge | Cycle throughput, charge acceptance, low-SoC time and energy efficiency |
| Diesel optimization | Controlled charge blocks and deeper battery discharge | Power, generator loading, start logic, fuel economics and recovery time |
| Weak-grid support | Frequent short outages, voltage disturbance or scheduled load shifting | Fast transfer, cycle count, protected reserve and rectifier coordination |
| Tariff or grid service | Scheduled dispatch while retaining telecom backup | Warranty boundaries, market rules, secure controls and reserve priority |
Ask one plain question: after the battery has performed its commercial or energy-saving task, how much reserve must remain for the network? That answer belongs in the controller logic and warranty model, not only in a sales presentation.
Evaluate the battery as a system
Cells are one layer of the product. A telecom battery system also includes busbars, module structure, a BMS, current sensing, balancing, contactors or breakers, branch fuses, communications, event memory and the cabinet or rack interface. The weakest of those layers can determine site availability.
A nominal 48 V label is not an interface specification. Some lithium iron phosphate products are marketed as 48 V class or 51.2 V nominal systems, but their permitted charge and discharge voltage depends on cell count, BMS settings and product design. Confirm all of the following against the rectifier and telecom load:
- Minimum and maximum battery terminal voltage
- Rectifier float, boost and current-limit setpoints
- Low-voltage disconnect and reconnect thresholds
- Permitted continuous, peak and charging current
- Pre-charge, contactor and wake-up behaviour
- Alarm interface, protocol version and safe fallback after communication loss
Parallel modules need particular attention. Packs should have compatible hardware and firmware, similar state of charge before connection, individual branch protection and a validated current-sharing strategy. A large state-of-charge difference can create an inrush or circulating current. Mixing battery models, ages or chemistries on a common bus should only be done under an architecture specifically designed and approved for it.
The ITU paper separates the BMS into data-acquisition and control functions: a battery management unit gathers cell voltage, current and temperature, while a higher control function applies charge, discharge, balancing and protection logic. Product names vary, but the design question is consistent: which protective action remains available if a sensor, processor, communication link or power supply fails? Remote alarms are helpful; an alarm that arrives after the battery has lost local protection is not.
| Initiating condition | Possible failure chain | Evidence or check |
|---|---|---|
| Loose terminal or poor crimp | Resistance rises, local heating develops, insulation or connection deteriorates | Connection design, torque control, end-of-line inspection and thermal check under load |
| Condensation, water or conductive debris | Leakage current, corrosion, tracking or short circuit | Enclosure strategy, drainage, conformal protection where applicable and site inspection |
| Cell inconsistency | One cell reaches a voltage limit first, usable string capacity falls, imbalance grows | Incoming-cell data, grouping method, voltage spread and capacity test |
| Sensor drift or BMS fault | Incorrect SoC or temperature decision, late alarm or unintended trip | Accuracy specification, plausibility checks, independent protection and fault-injection test |
| Mechanical shock or long-term cell expansion | Deformation, damaged insulation, leakage or connection stress | Mechanical restraint, clearance, transport packaging and inspection criteria |
This is why a battery incident should not be treated as a single-component story. A minor assembly defect can remain quiet for months, then meet high current, heat or moisture and become a service outage.
Size usable energy, power and end-of-life capacity
C_nominal = E_required ÷ (DoD × η_path × F_EOL × F_temp)E_required is the energy delivered to the telecom load during the autonomy interval. DoD is the allowable depth of discharge, η_path is the efficiency from battery to load, F_EOL is retained capacity at the specified end of life, and F_temp represents capacity available at the design temperature. Use manufacturer data for the selected product and avoid applying a derating factor twice if it is already included in a declared usable-energy figure.
For example, a constant 1.0 kW critical load requiring eight hours of autonomy needs 8.0 kWh delivered. With 80% allowable depth of discharge, 95% path efficiency, 80% retained capacity at end of life and a 0.90 cold-temperature capacity factor, the preliminary initial nameplate requirement is:
8.0 ÷ (0.80 × 0.95 × 0.80 × 0.90) = 14.6 kWhThe result must then be rounded up to a compatible number of modules. After rounding, recalculate current, allowable depth of discharge, recharge time and operation with one module unavailable if the site requires that level of redundancy.
Energy is only half the sizing task. Calculate current at the lowest permitted operating voltage, not merely at nominal voltage:
I_peak = P_peak ÷ (V_min × η_path)At 3 kW, 48 V and 95% efficiency, the screening current is about 65.8 A. The real worst-case value will be higher if the battery voltage falls below 48 V. Check BMS limits, contactors, busbars, fuses, cable ampacity and voltage drop at the actual minimum voltage and site temperature.
Recharge is a separate constraint. After an outage, the rectifier, PV array or generator must carry the live telecom load while restoring battery reserve. A large battery connected to a small source may meet the autonomy target but fail the required recovery time.
Temperature changes performance, life and control
High temperature generally accelerates ageing, while low temperature can reduce available discharge energy and restrict charging. There is no universal temperature-life multiplier that should be applied to every lithium battery. Use the cell and product manufacturer’s curves, test conditions and warranty limits.
Outdoor cabinet design should account for ambient extremes, solar gain, internal electrical losses, airflow restriction, dust, humidity, salt and altitude. Measure cell or module temperature as well as cabinet air. A cool air sensor near an inlet can hide a hot module deeper in the rack.
| Condition | Possible consequence | Design response |
|---|---|---|
| High ambient temperature and solar gain | Faster ageing, derated current and higher cooling load | Shade, insulation, ventilation or cooling sized from a thermal model |
| Charging below the permitted temperature | Cell damage or BMS charge inhibition | Charge lockout, controlled heating and an operating-state alarm |
| Large module-to-module temperature spread | Uneven ageing and state-of-charge estimation error | Improve airflow and trend multiple temperature points |
| Dust, salt or moisture ingress | Cooling loss, corrosion, tracking or connector failure | Suitable enclosure, filtration, drainage, seals and maintenance interval |
Heating and cooling consume energy. Include their duty cycle in the site load model; otherwise the battery cabinet can quietly invalidate the autonomy calculation it was intended to satisfy.
An IP code is useful, but it answers a limited enclosure question under defined test conditions. It does not, by itself, prove resistance to internal condensation, salt deposits, insects, metal swarf left during installation or ageing of cable glands. Match the enclosure and maintenance plan to the actual site. A sheltered indoor rack and a roadside outdoor cabinet should not inherit the same environmental assumptions.
Safety depends on independent layers
- Cell and module: cell traceability, manufacturing consistency, insulation, mechanical restraint and verified fault behaviour.
- BMS: cell-voltage and temperature monitoring, current limits, balancing, fault detection and a protective action that does not depend solely on remote software.
- Electrical protection: correctly rated fuses or breakers, isolation, protected conductors, selective coordination and a means to disconnect each parallel branch.
- Cabinet and site: thermal management, ingress and corrosion protection, detection where required, controlled access, separation and emergency instructions.
- Operations: approved firmware, alarm ownership, maintenance, transport, storage, replacement and end-of-life handling.
Chemistry matters, but labels such as LFP or NMC do not replace system evidence. Procurement teams should ask for reports that match the offered cell, module and battery configuration. IEC 62619:2022 covers safety requirements and tests for industrial secondary lithium cells and batteries, explicitly including stationary telecom applications. Its relevance does not remove the need to identify installation rules and approvals in the destination market.
The ITU paper groups useful validation work into mechanical, electrical, thermal, environmental and extreme-condition tests. That is a good way to inspect a supplier’s evidence, provided the test level is made clear.
| Test family | What it can investigate | What the buyer should verify |
|---|---|---|
| Mechanical | Drop, shock, vibration, crush or structural restraint | Whether the test was on a cell, module, rack or shipping package, and at what state of charge |
| Electrical | Overcharge, overdischarge, external short circuit, insulation and reverse connection | Fault duration, protective devices present and pass criteria |
| Thermal | High temperature, thermal shock, abnormal heating or propagation behaviour | Initiation method, sensor locations, observations and relevance to the installed configuration |
| Environmental | Water, humidity, salt mist, dust, corrosion, EMC or seismic stress | Severity level and whether it matches the destination site |
| Transport | Altitude simulation, vibration, shock, external short circuit and other transport hazards | Exact model coverage, test summary and shipment configuration |
A dramatic cell-level test video is not the same as a standards report for the delivered pack. Conversely, a pack certificate does not reveal every installation risk. Build a traceability chain from the offered model and cell to the report, factory configuration and final shipment.
Transport evidence is distinct from stationary product safety. The UN Manual of Tests and Criteria contains subsection 38.3 for lithium-battery transport testing. Buyers should request the applicable UN 38.3 test summary and verify that model numbers and battery configuration match the shipment. A safety data sheet alone is not a substitute for that evidence.
Treat SoC and SoH as estimates, not perfect measurements
State of charge is calculated from current integration, voltage, temperature and battery models; it is not measured directly like tank level. Sensor offset, incomplete calibration and long periods without a known reference point can cause drift. State of health is also model-dependent and may represent capacity, power capability, resistance or a combination of indicators.
Trend cell or module voltage spread, temperature, current, charge throughput, protection events, contactor operations and communication health. Periodic controlled capacity checks can test whether the displayed SoH agrees with delivered energy. Time synchronization matters: without consistent timestamps, a rectifier alarm, BMS trip and site outage are difficult to place in the correct order.
Monitoring should lead to action. Define thresholds for minimum reserve, high-temperature exposure, abnormal imbalance, repeated protective trips, unexpected self-discharge and reduced measured capacity. State of energy can be a more useful operating quantity than SoC when the controller must decide how many kWh remain for the load.
Balancing also needs context. Passive balancing dissipates excess energy as heat; active balancing transfers energy between cells and can improve capacity use. The ITU paper favours active balancing for higher efficiency. That should not become a checkbox in isolation: active circuits add components and control complexity, while a well-matched, lightly cycled backup string may have different needs from a daily-cycled solar site. Ask for balancing current, activation conditions, thermal effect, fault behaviour and evidence from the intended duty cycle.
ITU-T L.1397 provides a current telecom-specific information model for integrated battery units and systems. Whether or not that exact interface is adopted, agree on parameter names, units, alarm severity, sampling interval, timestamp source, data retention and behaviour after communication loss before commissioning.
Remote access adds an operational security requirement. Use unique credentials, role-based permissions, protected management channels, network segmentation, controlled firmware updates, audit logs and a documented local fallback. Loss of the cloud connection should not prevent basic battery protection or the site’s agreed backup function.
Buy evidence, not only specifications
| Supplier claim | Evidence to request | Question to settle |
|---|---|---|
| Rated and usable capacity | Discharge curves at stated rate and temperature | How many kWh reach the load at beginning and end of life? |
| Long cycle life | Test conditions and warranty schedule | At what DoD, temperature, current and end-of-life threshold? |
| High power | Continuous and peak-current duration curves | Can every protective component carry the site peak? |
| Safe parallel operation | Approved topology, current-sharing data and fault response | What happens if one module or communication link fails? |
| Manufacturing consistency | Incoming inspection, cell grouping, ageing test and serialized production records | Can a field unit be traced to cell batch, firmware and end-of-line results? |
| Outdoor suitability | Environmental reports, enclosure drawings and installation limits | How are condensation, insects, corrosion and cable entries managed? |
| Standards compliance | Certificates and reports traceable to the offered model | Do cell, module and system configurations match? |
| Remote monitoring | Protocol map, alarm list, access model and firmware policy | Who controls updates and what works when communications fail? |
| Service support | Spare-parts plan, response process and version policy | Can a failed module be replaced without destabilizing the bank? |
| Transport readiness | UN 38.3 test summary and shipping documentation | Is the exact shipment configuration covered? |
Read the warranty alongside the datasheet. Cycle count may be conditional on depth of discharge, temperature, C-rate, throughput, state-of-charge window and communication with an approved controller. Calendar expiry may occur before the advertised cycle number on a lightly used backup site.
Factory quality records matter because several credible failure initiators are mundane: a loose fastener, foreign metallic material, a poorly terminated cable or a cell group with excessive spread. Useful evidence includes incoming inspection, cell-capacity grading, insulation and hipot results where applicable, end-of-line functional tests, ageing records, torque or joining-process controls and serialized links between the product and its test data. A generic ISO 9001 certificate does not answer whether those controls were applied to the offered battery.
Site acceptance should confirm model and serial numbers, mechanical condition, branch protection, polarity, torque records, insulation, communication, alarm mapping and controller setpoints. A controlled load test and recharge observation provide stronger evidence than a dashboard showing 100% SoC. Record the as-built firmware and configuration so a later service visit has a reliable baseline.
Storage and delivery deserve their own procedure. Use the manufacturer’s state-of-charge, temperature, humidity and refresh-charge limits rather than copying a generic warehouse range. Check packaging after transport, quarantine any pack with impact, deformation, leakage or abnormal voltage, and control the power-on sequence during installation.
For European projects, communication-infrastructure batteries generally fall within the industrial-battery category under Regulation (EU) 2023/1542. The current regulation introduces lifecycle, information and end-of-life obligations in stages. From 18 February 2027, each industrial battery above 2 kWh placed on the EU market or put into service is scheduled to require a battery passport. Importers and distributors should verify which obligations apply to their role and product at the actual placing-on-market date rather than treating a CE mark as the whole compliance file.
End-of-life planning should identify who owns removal, safe packaging, transport and recycling, and what operating data will accompany the battery. That information affects residual-value decisions and makes a damaged or aged pack less ambiguous to handle.
Huijue can configure telecom battery systems around a defined duty cycle and interface specification. Final product selection should still be based on approved site data, destination-market requirements and evidence for the exact configuration being supplied.
What should you send for a telecom battery proposal?
Provide the hourly load or measured average and peak demand, autonomy target, DC voltage window, charging sources, site temperature range, installation arrangement, communication protocol and expected operating duty.
Send these requirements through your established Huijue sales or engineering contact.