Liquid-Cooled BESS Maintenance Checklist for Operators

Published: September 30, 2026 | Author: Weltrus Energy Team | Reading Time: 14 minutes

Scope: O&M After the Plant Is Live

Commercial and industrial liquid-cooled battery energy storage systems (BESS) shift failure modes from “fans clogged” toward loop integrity, coolant quality, and heat rejection. Operators who only copy air-cooled filter routines miss pump cavitation, glycol degradation, and exchanger fouling until cell temperatures drift and the BMS derates or trips.

This article is an operator maintenance checklist. It is not the topology pitch in why liquid-cooled BESS, not the liquid vs air C&I comparison, and not the noise and installation guide. Commissioning and factory acceptance sit upstream in grid-tied solar/BESS commissioning and BESS FAT before shipment. Fleet dashboards are covered in SCADA fleet monitoring.

Always overlay OEM intervals, fluid type, and PPE. Where this list conflicts with the manufacturer’s O&M book, the book wins. Print the final site version with revision date next to the cooling skid so contractors do not pull an obsolete PDF from email.

Shift / Daily SCADA and Walk Checks

1. Loop supply / return temperatures — Compare to baseline at similar ambient and SOC; note rising delta-T with stable load.
2. Module or pack max/min temperature — Flag spreads outside OEM limits even if the average looks fine.
3. Coolant pump status and alarms — Run/stop, speed if VFD, trip history, and “pump not running while cooling demand active.”
4. Pressure (or equivalent) — Low pressure can mean leak or air; high pressure can mean blockage or closed valve.
5. Leak / humidity / door alarms — Clear nuisance tags only after a physical check.
6. BMS thermal derate or power limit events — Log start/stop times; pattern often points to heat rejection, not cells alone.
7. Quick exterior walk — Visible drip trays, puddles under cabinets, frost or wet insulation on outdoor lines (site-dependent).

Record ambient dry-bulb (and wet-bulb if the dry cooler depends on it). Thermal complaints without ambient context waste OEM support time.

Weekly Physical Checks

• Inspect hose, quick-connect, and flange areas for residue, staining, or damp dust.
• Confirm isolation valves are in the expected open/closed state after any work order.
• Listen for unusual pump noise (cavitation, bearing growl) during a commanded cool cycle if safe.
• Check cabinet and cooler intake paths for debris; liquid-cooled plants still need clear air across dry coolers or condensers.
• Verify emergency stops and remote trip paths were not defeated during the prior week’s work.
• Photo any new corrosion at fittings; coastal and high-pollution sites accelerate this.

Do not open pressurized loops without lockout and OEM procedure. “Just cracking a fitting to check” is how glycol ends up in cable trays.

Monthly Loop and Filter Tasks

If the site uses a water–glycol mix, treat concentration and inhibitor life as measured values, not tribal knowledge. Wrong mix shows up first as freeze risk or corrosion, then as warranty arguments.

Annual / Interval Deep Service

1. Coolant sample — Lab or field kit per OEM (conductivity, pH, inhibitor, particulates, biological growth if water-based).
2. Pump and seal service — Hours-based rebuild or replacement; align coupling if present.
3. Heat exchanger clean — Chemical or mechanical clean as specified; document before/after approach temperatures.
4. Valve exercise — Stroke isolation and control valves; confirm feedback to BMS/EMS.
5. Electrical on cooling skid — Torque check on power terminals, contactor wear, VFD cooling fans, ground continuity.
6. Calibration — Temperature, pressure, and flow sensors used for protection trips.
7. Spare kit audit — Seals, clamps, approved fluid, filters, one spare pump or seal kit if the OEM recommends it.

Schedule deep service with a planned outage window. Partial bypass without OEM approval can leave packs unprotected while the skid is offline.

After fluid work, document a return-to-service proof: stable loop pressure for a defined hold time, no new alarms during a controlled discharge/charge thermal cycle, and temperatures converging toward the site baseline. Skip that proof and the next shift inherits an unverified loop.

CMMS Fields Worth Keeping

• Asset ID (cabinet / string / cooling skid) and software/firmware revision of BMS thermal package
• Fluid specification (brand/type), batch or drum ID, fill and top-up volumes with date
• Sample results and next sample due date
• Pump runtime hours and filter change dates
• Thermal derate events: duration, max cell temp, ambient, corrective action
• Leak events: location, volume estimate, cleanup, retest pressure
• Work order photos before close-out

Fleet owners comparing multiple sites need the same field names. Inconsistent tags make SCADA trends useless when support asks “what was ΔT last August.”

When to Stop and Escalate

• Visible coolant leak into electrical compartments or pooling under HV DC sections
• Rapid unexplained drop in loop pressure or level
• Persistent thermal derate after cooler and filter service
• Pump trips that recur after reset
• Sensor disagreement that would mask a real over-temp
• Any odor of glycol or burnt insulation near the skid

Isolate per site LOTO, notify the OEM or qualified service partner, and keep the CMMS ticket open with raw SCADA exports attached—not a one-line “cooling issue.” Attach the last filter change date and the latest coolant sample result when you open the ticket; those two fields cut days off remote triage.

Boundaries

• Not a substitute for the OEM O&M manual or local electrical code.
• Not fire-suppression or BMS logic design guidance (see related fire/BMS specs content on Weltrus where needed).
• Not a claim that liquid cooling needs zero maintenance.
• Not permission to mix coolant chemistries across cabinet generations.

Operators who keep loop health visible usually catch issues before cycle-life arguments start. The checklist above is the minimum structure; fill intervals from your OEM book, freeze them in the CMMS, and review missed work orders in the weekly operations meeting.

Frequently Asked Questions