Auxiliary & Backup Power Supply for ESS Systems
← Back to: Energy & Energy Storage Systems
Auxiliary and backup PSUs in ESS cabinets are the hidden power plant that keeps BMS, fire protection, sensing and gateways alive when the main PCS path trips, surges or shuts down. Proper isolation, hold-up, eFuse protection and PG sequencing turn this network from a fragile add-on into a safety-critical backbone.
What this page solves
An energy storage system is usually designed around the main power path: battery packs, power conversion system (PCS) and inverters that push energy to the grid or to local loads. However, safe operation and observability depend on a separate auxiliary and backup power supply that keeps control, sensing and safety functions alive even when the main path is shut down or unavailable.
During faults, blackouts, fire events or intentional isolation, the PCS and main DC bus may be disconnected by design. Critical subsystems such as pack and module BMS, insulation monitors, gas and thermal runaway sensing, fire detection and suppression modules, and the site gateway or EMS still need reliable low-voltage rails to record data, complete orderly shutdown sequences and trigger protective actions.
This page focuses on the auxiliary and backup PSU as a small, isolated and safety-grade power plant inside an ESS. The goal is to provide stable, low-noise and well-sequenced supply rails, with proper isolation, hold-up capability, power-good signalling and protection, instead of using ad-hoc DC/DC modules tied directly to the main power stage.
The scope is limited to auxiliary and backup power for control, sensing and safety circuits. Main PCS and inverter power stages, buffer ESS for fast charging, UPS battery systems and supercapacitor energy modules are covered on their dedicated pages in this energy and storage system cluster.
ESS subsystems that depend on auxiliary and backup power
Inside an ESS, auxiliary and backup rails do not serve a single board. They feed several critical subsystems with very different power profiles, noise tolerance and safety roles. The table below highlights the most important consumers, the type of supply they require and the risk if the aux or backup PSU is poorly designed or lost during a disturbance.
| Subsystem | Supply requirements | Risk if supply is lost |
|---|---|---|
| Pack / module BMS | Stable 5 V / 12 V rails with low noise and controlled sequencing for ADCs, AFEs and MCUs. | Loss of cell voltage and temperature visibility, incomplete contactor actions and missing fault logs. |
| Fire detection & suppression interface | Long hold-up capability for smoke sensing AFEs and short, high-current pulses for igniters or valves. | Fire events may be detected too late or suppression may not complete because supply collapses mid-action. |
| Thermal runaway / gas sensing | µA to mA bias currents from low-noise rails for gas, pressure and optical SoCs plus alarm drivers. | Higher false-alarm rate or missed early TR signatures due to noisy, unstable or interrupted supply rails. |
| Gateway / EMS / communications | Sequenced multi-rail supply for secure boot, SoCs, modems and switches, with local hold-up for log flush. | Repeated brown-outs, corrupted event records and gaps in fleet-level monitoring during grid or PCS trips. |
| Insulation and leakage monitor | Isolated, low-noise supply for injection sources, measurement AFEs and digital processing blocks. | Loss of real-time visibility of HV-to-ground insulation, reducing the ability to detect evolving faults. |
These subsystems collectively cover safety actions, safety sensing and system visibility. Because they rely on auxiliary and backup PSU rails instead of the main PCS path, the design of those rails must be treated as safety-grade infrastructure rather than a convenience add-on.
Typical architectures & power profiles
Auxiliary and backup PSUs can be grouped into three broad architectures based on how they operate over time and how they interact with the rest of the ESS: always-on power supplies that keep monitoring alive, emergency backup paths that support short but demanding actions, and cold backup paths that only wake up under specific fault or maintenance conditions.
Always-on power typically uses flyback or buck controllers with post-regulation rails to supply BMS controllers, critical sensors and low-power communication logic continuously. Emergency backup stages often combine LLC or flyback converters with dedicated hold-up capacitors or supercapacitors to deliver several seconds to minutes of energy for fire suppression, contactor actuation or log flushing after a PCS trip. Cold backup rails remain off until a defined event triggers an eFuse or high-side switch and a sequencing circuit, providing a controlled wake-up path for diagnostics, redundant controllers or service ports.
This page focuses on the topology and power-path concepts for these three classes of auxiliary and backup supplies. Supercapacitor balancing, lifetime management and dedicated fast-charging buffer ESS designs are covered on separate supercapacitor and buffer ESS pages within this energy and storage system cluster.
Design requirements & constraints
Auxiliary and backup PSUs inside an energy storage system must be specified and designed as safety-grade components rather than generic catalog AC/DC modules. Isolation, standby losses, electromagnetic robustness, protection behaviour, output noise, power-good timing, firmware update support and thermal performance all become system-level constraints that link directly to functional safety, diagnostics and fleet availability.
Reinforced isolation. Isolation between the high-voltage battery bus and low-voltage control domains must satisfy reinforced insulation requirements under standards such as UL and IEC 61010 or IEC 61508. This includes adequate creepage and clearance around transformers and high-voltage nodes, coordination with insulation monitoring functions and clear separation between different communication grounds when several isolated interfaces share the same auxiliary supply.
Low standby consumption. Many ESS deployments operate continuously over years across dozens or hundreds of cabinets. The quiescent current of auxiliary controllers, eFuses, sequencing ICs and bias rails accumulates into a fleet-level energy cost. Always-on rails therefore need converters and regulators that maintain high efficiency at light load, while still delivering the low noise required by measurement chains.
EMI and surge immunity. Auxiliary supplies sit in the same cabinet as PCS and inverters, and share long harnesses that are exposed to surge, common-mode noise and fast transients. Input filtering, surge protection and proper layout are required so that the aux or backup rail remains operational through EMC testing and realistic grid events, without injecting noise into BMS ADCs or insulation monitors that depend on low-distortion signals.
Protection: OCP, OVP and OTP. Over-current, over-voltage and over-temperature protection must guard not only the power converter itself, but also downstream loads and the upstream battery or DC bus. Current-limited eFuses, programmable trip profiles and carefully chosen shutdown thresholds are required to distinguish between legitimate high-current actions, such as fire suppression triggers, and true faults that should disconnect the rail.
Output noise for measurement and references. Many safety-related algorithms rely on precise measurements of cell voltage, temperature, gas concentration and insulation impedance. Auxiliary rails that feed AFEs, ADCs and reference circuits must therefore be treated as precision supplies, with appropriate post-regulation, filtering and grounding to keep noise and ripple within the limits assumed in BMS and diagnostics calculations.
Power-good sequencing and timing. Start-up and shutdown sequences across multiple rails influence contactor logic, secure boot flows and logging behaviour. Power-good signals and supervisors must enforce stable ramp-up for cores, I/Os and communication interfaces, and provide early warning on voltage decay so that BMS, gateways and safety controllers can complete contactor actions and write critical data before a rail falls out of regulation.
Firmware update and OTA support. Many ESS subsystems now update firmware remotely, including BMS controllers, gateways and safety interfaces. Backup and auxiliary supplies must support enough energy and runtime to complete secure download, verification and programming cycles without interruption, and must cooperate with dual-image or rollback schemes to avoid rendering devices inoperable after a transient or brown-out.
Thermal dissipation in sealed cabinets. Auxiliary PSUs often operate in sealed or semi-sealed enclosures with limited airflow and elevated ambient temperatures. Component derating, placement relative to high-loss PCS hardware and realistic thermal modelling are required to ensure long-term reliability. Protection thresholds and thermal design must accommodate hot-day operation without nuisance trips, while still preventing overstress during rare but severe overloads.
IC categories & selection checklist
Auxiliary and backup PSUs in ESS cabinets rely on a small set of IC categories that repeat across designs: primary-side controllers for flyback or LLC stages, synchronous rectifier drivers, eFuses and high-side switches, power-good and sequencing controllers, and supervisor or reset ICs. The table below focuses on category-level applicability and key parameters, without naming specific vendors.
| IC category | Typical role | Key selection metrics |
|---|---|---|
| Flyback / buck controller | Base isolated or non-isolated supply for always-on rails and medium-power auxiliary outputs. | Light-load efficiency, switching strategy, noise performance, start-up behaviour and bias consumption. |
| LLC resonant controller | High-efficiency stages with sizeable hold-up energy for emergency backup rails. | Soft-start control, light-load and burst mode behaviour, frequency range and protection features. |
| Synchronous rectifier driver | Efficiency improvement on secondary sides of flyback or LLC stages across a wide load range. | Valley switching support, adaptive timing, low reverse conduction and stable operation at light load. |
| eFuse / high-side switch | Branch protection and controlled inrush for BMS, fire, gateway and diagnostics rails. | Adjustable ILIM, programmable trip curves, telemetry, reverse blocking and thermal shutdown behaviour. |
| PG / sequence controller | Coordination of rail ramp-up and ramp-down for BMS, gateway and safety controllers. | Configurable delays, threshold accuracy, active enable control and fault handling modes. |
| Supervisor / reset IC | Undervoltage and brown-out protection for MCUs, AFEs and communication SoCs. | Precise reset threshold, hysteresis, reset delay, wide temperature range and low quiescent current. |
A robust auxiliary and backup PSU design usually combines several of these IC categories: flyback or LLC controllers set the power stage behaviour, synchronous rectifier drivers recover efficiency, eFuses and high-side switches define branch-level protection and diagnostics, while sequencing and supervisor ICs enforce the power-on and power-down rules assumed by BMS and gateway firmware.
Application mini-stories
Real projects often expose auxiliary and backup PSU weaknesses only after field operation. The following short application stories illustrate how the IC categories above combine to solve practical issues in containerised ESS deployments, emergency hold-up paths and fire suppression rails.
Container ESS gateway resets — aux rail as the hidden culprit
In one container ESS, the site gateway and EMS controller were fed directly from a PCS auxiliary output without a dedicated aux rail or sequencing logic. Every time the PCS tripped on a grid disturbance, the auxiliary node sagged, the gateway experienced brown-out resets and secure boot restarted. Event logs around the fault were incomplete, leaving operation teams with reset entries but no clear root cause.
A revised design introduced U1 as a flyback-based always-on converter, U2 as an eFuse on the gateway branch, U5 as a PG and sequence controller and U6 as a supervisor for the gateway SoC. U1 now decouples the aux rail from PCS trips, U2 limits inrush and records branch faults, U5 enforces the correct rail ramp-up order and U6 ensures a clean reset before the SoC supply drifts below specification. After the change, gateway logs consistently capture PCS events and shutdown decisions instead of showing unexplained resets.
Supercap standby — sizing for a 15 second EMS hold-up
An ESS design required the EMS controller to remain online for at least 15 seconds after a DC bus trip so that logs could be flushed and a status summary uploaded. The initial implementation relied on a small auxiliary flyback with limited output capacitance. In practice, the EMS rail dropped out after less than a second under full load, interrupting log writes and leaving diagnostics incomplete.
The updated design used U3 as an LLC resonant controller feeding a dedicated emergency backup rail, Chold sized from the EMS power profile and hold-up requirement, and U4 as a synchronous rectifier driver to minimise secondary losses. An eFuse on the EMS branch limits inrush when the rail is recharged and reports overload conditions. With the combination of higher-efficiency conversion and correctly dimensioned supercap storage, the EMS now reliably meets the 15 second hold-up target with thermal margins in a sealed cabinet.
Fire suppression rail — transient pulses and disciplined protection
A fire detection and suppression module in another ESS cabinet shared its supply with BMS and thermal runaway sensing circuits. When igniters and valves were driven during tests, large transient pulses on the shared rail caused voltage dips that disturbed the measurement chain and occasionally triggered nuisance resets in nearby controllers. The supply had no dedicated branch protection, and rectification relied on diodes with significant conduction loss.
A separation of rails and IC roles resolved the issue. U1 provided a dedicated emergency rail for fire actuation, U4 drove synchronous rectifiers to keep droop small under pulse load, and U2a implemented an eFuse on the fire branch with tuned current limits and thermal behaviour. BMS and sensing modules were moved to their own branches behind U2b, an additional eFuse with a lower ILIM. Sequencing logic coordinates which rails may temporarily droop during a suppression event and which must remain within tight limits. As a result, suppression pulses no longer disturb measurement accuracy or safety controllers elsewhere in the cabinet.
Recommended IC roles mapping
Inside an ESS cabinet, a central auxiliary and backup PSU feeds multiple functional blocks that each require specific IC roles for conversion, protection, sequencing and supervision. The mapping below links key ESS subsystems to the type of auxiliary rail they depend on and the typical IC roles that appear in their schematics, expressed as designators and categories rather than vendor-specific part numbers.
| ESS functional block | Aux / backup rail type | Key IC roles (designator & category) | Purpose in ESS |
|---|---|---|---|
| Pack / rack BMS controller | Always-on rail for safety-critical monitoring | U1: flyback / buck controller; U2: eFuse for BMS branch; U5: PG / sequence controller; U6: supervisor IC for BMS MCU. | Keeps the BMS powered through PCS trips, enforces rail sequencing for AFEs and MCUs and protects the BMS branch against shorts and brown-outs. |
| Module BMU / CMU and cell AFEs | Always-on or duty-cycled low-power aux rail | U1m: isolated flyback controller; U2m: high-side switch for module rail; U6m: supervisor for local MCU or AFE digital core. | Supplies accurate, low-noise power to cell monitoring AFEs while allowing selective shutdown of modules for service or storage. |
| Insulation & leakage monitor | Always-on isolated rail with tight noise limits | U1i: flyback controller with reinforced isolation; U2i: eFuse to separate injection and measurement circuits; U6i: supervisor for insulation MCU. | Maintains clean, low-distortion supplies for injection and AFE stages so that insulation impedance readings remain reliable under grid disturbances. |
| TR / gas / smoke sensing cluster | Always-on low-noise sensor rail | U1s: bias rail regulator fed from aux; U2s: eFuse per sensor group; U6s: supervisor or window detector for AFE reference rails. | Provides stable, low-ripple bias for NTC, RTD and gas AFEs while allowing faults in one sensor group to be isolated without losing the entire cluster. |
| Fire detection & suppression controller | Emergency backup rail for pulsed high current | U3f: LLC controller for actuation rail; U4f: synchronous rectifier driver; U2a: eFuse on fire branch; U5f: PG coordination with BMS and gateway rails. | Delivers short, high-current pulses to igniters and valves without collapsing other control rails, and records branch faults for maintenance analysis. |
| LED / buzzer / local alarm interface | Always-on rail with limited current capability | U1a: small buck or LDO from aux rail; U2a2: eFuse or load switch; Q1–Qn: low-side drivers for indicators and buzzers. | Keeps visual and acoustic alarms powered even when main control logic is restarting, while limiting fault current on long panel harnesses. |
| Site gateway / ESS EMS controller | Always-on rail with emergency hold-up extension | U1g: flyback controller; U3g: LLC controller with Chold for N-second hold-up; U2g: eFuse for gateway rails; U5g: PG for SoC rails; U6g: supervisor. | Enables reliable secure boot, log flushing and status reporting during grid faults by guaranteeing a controlled supply and reset profile for the gateway. |
| Secure OTA / firmware update paths | Always-on or emergency-backed rail for update window | U1o: aux converter feeding OTA controller; U2o: eFuse to isolate update path; U5o: PG logic to guarantee monotonic ramps during programming. | Ensures that firmware download, verification and programming sequences are not interrupted by rail glitches, reducing the risk of bricked controllers. |
| Cabinet / container environment monitor | Low-power always-on rail, possibly cold-start capable | U1e: flyback or buck controller; U2e: high-side switch per sensor bus; U6e: supervisor for low-power MCU; Qe: drivers for door / leak alarms. | Keeps temperature, humidity, leak and door sensing alive to support early detection of cabinet issues, even when main PCS is disabled. |
| DC bus & ground fault localisation | Always-on or emergency-backed diagnostic rail | U1d: isolated aux converter; U2d: eFuse for distributed probes; U6d: supervisor for correlation MCU; optional PG link to BMS for coordinated sampling. | Maintains the ability to localise faults on the DC bus and grounding network during abnormal events, supporting faster root-cause analysis. |
| Cold backup / service controller | Cold backup rail with wake-on-fault capability | U1c: charger for backup battery or supercap; U2c: eFuse / HS switch controlling cold rail; U5c: PG logic for controlled wake; U6c: supervisor for service MCU. | Allows a minimal controller to wake only when diagnostics or service are required, limiting standby losses while preserving fault visibility. |
Treating these IC roles as reusable building blocks makes it easier to keep auxiliary and backup power designs consistent across multiple ESS products. Each new project can start from a small set of proven controller, rectifier, eFuse, sequencing and supervisor building blocks mapped to the same functional tree.
Central aux / backup power tree inside an ESS
The power tree below summarises how a central auxiliary and backup PSU taps the main DC bus and fans out always-on, emergency and cold backup rails to pack BMS, fire suppression, gateway and EMS controllers, TR and gas sensing, OTA paths and local alarm interfaces. Branch eFuses, sequencers and supervisors sit along the tree to enforce protection and timing assumptions used throughout the design.
Request a Quote
FAQs for auxiliary and backup PSU in ESS
The questions below highlight typical design trade-offs around auxiliary and backup PSUs in energy storage systems: when to split rails, how to coordinate power-good sequencing, which IC features matter for emergency hold-up, how to set eFuse limits and how to avoid recurring field failure patterns.
