Emergency Power ATS Monthly Load-Transfer Test & Battery Condition Log

ContextFloat voltage is the steady-state voltage the charger maintains to keep a fully charged battery ready for service. For a standard 12V VRLA (valve-regulated lead-acid) battery, a healthy float is between 13.2 V and 13.8 V per unit; flooded lead-acid batteries float slightly lower. Record individual readings, not a single aggregate reading for the bank—one low cell masks itself in an average. A battery reading below 12.8 V at float has likely lost significant capacity and will not deliver rated cranking amperage. Replace any unit measuring below 12.4 V immediately; a single weak unit drags down the rest of the string under load.

ContextSpecific gravity measures the concentration of sulfuric acid in the electrolyte—a direct proxy for state of charge and overall battery health. A fully charged cell reads 1.265–1.280 at 77°F (25°C). Readings below 1.220 indicate undercharge or sulfation. A spread of more than 0.030 between the highest and lowest cell in a bank indicates an imbalanced string that will fail under load even if average readings look acceptable. Always temperature-compensate your reading: add 0.004 for every 10°F above 77°F, subtract 0.004 for every 10°F below. Skip this step entirely for sealed VRLA or AGM batteries—puncturing them causes irreversible damage.

ContextTerminal corrosion (white or blue-green powder) increases resistance and can reduce cranking voltage by 10–20% even in a fully charged battery—enough to prevent a clean engine start in cold weather. Use a wire brush and a baking soda solution (one tablespoon per cup of water) to neutralize and remove corrosion, then rinse with distilled water and dry thoroughly before retightening. Check torque against the battery manufacturer's specification, usually 50–70 inch-pounds for standard post terminals; both over- and under-torquing cause damage. After cleaning, apply a thin coat of anti-corrosion terminal grease or petroleum jelly to the cleaned contacts.

ContextBattery capacity is highly temperature-sensitive. A lead-acid battery loses roughly 1% of capacity for every 1°F drop below 77°F (25°C), meaning a battery in a 40°F mechanical room may only deliver approximately 63% of its rated capacity. Conversely, sustained temperatures above 95°F accelerate sulfation and cut battery lifespan roughly in half for every 15°F rise above 77°F. If the battery room temperature is outside the 65–85°F range, flag it as a facilities infrastructure issue—correcting an HVAC or ventilation problem is almost always cheaper than replacing batteries ahead of schedule.

ContextExposed lead plates above the electrolyte surface sulfate rapidly and permanently lose capacity—damage that no amount of charging can reverse. The correct fill level is typically 1/8 inch below the bottom of the fill tube or vent well, never above it; overfilling causes electrolyte to spew from vent caps during charging, creating a corrosive mess and a safety hazard. Use only distilled or deionized water—tap water introduces minerals that accelerate plate degradation. Never add sulfuric acid (electrolyte) to a battery unless you are restoring one that had electrolyte physically spilled from it.

ContextA conductance tester (such as the Midtronics MDX series or equivalent) gives a fast, non-destructive indication of battery capacity in cold cranking amps (CCA). A full carbon-pile load test draws actual rated CCA for 15 seconds and measures voltage hold—more definitive but causes slight battery wear. Either method is acceptable; the critical metric is whether the battery maintains voltage above the manufacturer's minimum (usually 9.6 V for a 12V battery under full CCA load) for the full test duration. Any battery testing below 80% of rated CCA should be flagged for replacement planning within the next 90 days.

ContextA swollen or bulging battery case indicates thermal runaway or overcharge—an active safety hazard requiring immediate action. Isolate and remove a swollen battery immediately; do not attempt to discharge or recharge it. Brown or black discoloration on the case top often signals chronic overcharging from a faulty charger float voltage setting—a charger calibration issue, not a battery issue. Electrolyte seepage (a brown or crystalline residue around vent caps or seams) indicates either overfilling or excessive gassing. Photograph any abnormal findings for the equipment log and submit a corrective work order before leaving the room.

ContextDocumenting normal-source conditions before the test establishes a clean reference point. If utility voltage is already outside nominal (±5% of 480V or 120/208V depending on system) before the test begins, note it—your ATS may not transfer cleanly back at test conclusion. A phase imbalance exceeding 2% on the utility before transfer can also mask itself as an ATS problem in post-test analysis. Record all three phase-to-phase and phase-to-neutral voltages plus frequency, using a calibrated meter or the ATS's built-in metering display.

ContextThree common methods exist for initiating a monthly ATS test: (1) pressing the 'Test' button on the ATS controller, which simulates a utility-loss signal without actually dropping the utility—safest for occupied buildings; (2) opening the main utility breaker in a controlled sequence—more realistic but requires coordination with the electrical engineer; (3) using a remote test signal from the building automation system. The Test button method is the most common for monthly exercises but does not exercise the ATS under a true voltage dropout, which can mask certain timing relay faults. Document the method used each month so trend comparisons between months are meaningful.

ContextThis interval is set by the ATS's time-delay-on-start (TDNE) relay—typically 1–10 seconds for commercial applications and up to 30 seconds for some industrial settings. The delay prevents nuisance generator starts during momentary utility blips. If the observed delay deviates more than ±0.5 seconds from the set point (usually labeled inside the ATS door), the timing relay needs calibration or replacement. Use a stopwatch or the ATS event-log timestamp for measurement, not a smartphone—cell phone reaction time introduces 0.3–0.5 second of human error into the reading.

ContextMost diesel generators are configured to achieve rated output—within ±10% of rated voltage and ±0.5 Hz of rated frequency—within 10–15 seconds of receiving the start command. NFPA 110 Type 10 systems (hospitals, data centers, emergency egress lighting) require the generator to reach rated conditions within 10 seconds of utility loss. If your generator exceeds 15 seconds, investigate the cold-start fuel system, including the pre-heat system, injection timing, and block heater. Record both the crank-to-fire time and the fire-to-stable-output time separately—they diagnose fundamentally different problems.

ContextOnce the generator achieves stable output, the ATS's time-delay-on-transfer (TDEN) relay adds a brief additional delay—typically 1–5 seconds—before physically switching the load. This prevents connecting load to a generator that is technically at rated voltage but still hunting for frequency stability. The total observable event—from test initiation to the moment building loads are visibly energized from generator power—is the metric occupants experience. For NFPA 110 Type 10 systems, this complete sequence must finish within 10 seconds of utility loss. Time it end-to-end, not just segment by segment, and record both.

ContextThese four readings together confirm the generator is performing correctly under your actual building load. Output voltage should be within ±5% of rated. Frequency should hold within ±0.5 Hz. Power factor on a typical resistive and mixed commercial load runs 0.8–0.9 lagging; a power factor below 0.7 may indicate a highly inductive load condition stressing the alternator. Real load in kW is the most operationally critical: if building load exceeds 70–75% of generator rated capacity, the engine may struggle to regulate frequency during load steps, risking instability in sensitive downstream equipment.

ContextMost microprocessor-based ATS controllers maintain an internal timestamped event log. Even if no fault light stays on after the transfer, scroll through the event history menu and photograph any faults that appeared and cleared during the sequence—'glitch' faults that self-clear are early warnings of failing relays, contactor wear, or firmware instability. Common ones to document: contactor position disagreement, overcurrent on transfer, engine fault codes (low oil pressure at startup, high coolant temperature), and communication bus errors between the ATS and the generator control module.

ContextGenerator power confirmed at the ATS breaker does not mean every downstream load is functioning. Do a fast confirmation with each critical area: server room status lights, elevator recall panels, emergency egress lighting, ICU and OR power panels, fire alarm control panels, and any life-safety systems. A tripped sub-panel breaker or a UPS that failed to synchronize can leave a critical area dark even when the ATS transferred perfectly. This two-minute check is your protection against a test that shows a perfect transfer record while someone in the building was without power.

ContextThese three parameters track engine health during a loaded run. Oil pressure should stabilize within 10–15 seconds of startup and hold steady throughout the run; pressure that creeps downward over 20–30 minutes suggests a weak oil pump or developing bearing wear—a serious finding. Coolant temperature should rise steadily and stabilize at the thermostat set point (typically 185–195°F for diesel engines); if it continues rising past 210°F, initiate a controlled shutdown immediately. Exhaust temperature measured with a non-contact pyrometer at the turbo outlet gives a quick combustion efficiency read without any invasive access.

ContextThe 30-minute, 30%-load minimum is codified in NFPA 110 Section 8.4 for Level 1 and Level 2 Emergency Power Supply Systems. Running below this threshold allows wet-stacking: unburned fuel and carbon accumulate in the exhaust system because the engine never reaches full operating temperature, gradually fouling cylinder walls, piston rings, and turbochargers. Wet-stacked engines eventually smoke black under load, lose power output, and require either a regeneration run or expensive internal engine work. If your building load naturally falls below 30% during off-hours testing, you must supplement with a portable load bank—rental cost is typically $150–$400 per day depending on rated kW.

ContextTracking fuel consumption per test creates a trend line that detects injector wear, turbocharger efficiency loss, or building load changes before they become failures. A diesel generator's typical consumption is approximately 0.07 gallons per kWh at rated load—a 100 kW generator carrying 70 kW of actual load should burn roughly 4.9 gallons per hour. If consumption climbs more than 15% above the historical average without a corresponding increase in measured load, flag it for the engine service technician. Also note the absolute post-test fuel level: if the tank is below 50% after the run, schedule refueling before the next test date.

ContextMost ATS units incorporate a time-delay-on-normal (TDNR) relay that holds the load on generator for 0–30 seconds after utility power is confirmed stable, preventing premature retransfer to a source that may not yet be fully regulated. Open-transition ATS units will momentarily drop load during the switch (typically 1–4 cycles, or 16–67 milliseconds)—this is expected and normal. Closed-transition (make-before-break) ATS units briefly parallel utility and generator; ensure frequency and phase synchronization is confirmed before closed-transition occurs, or you risk a severe mechanical shock to the generator. Record the observed retransfer time and any visible interruption to non-UPS-protected loads.

ContextAfter retransfer, the generator should automatically enter a cool-down run—typically 3–10 minutes at no load—before shutting down. This slow cooling protects the turbocharger: an abrupt hot shutdown causes oil to 'coke' in the turbo bearing housing, which is the leading cause of premature turbocharger failure (replacement cost: $1,500–$8,000 depending on engine size). Do not manually stop the engine during cool-down. Once the engine shuts down, the battery charger should resume; allow 15 minutes and then verify charger output current is falling toward its float maintenance level, confirming the charger is functioning normally.

ContextA post-run inspection catches problems that only appear under thermal load: new oil seeps at gaskets or valve covers, coolant drips from hose connections, fuel odors near injectors or the fuel filter housing, and heat-blued metal near the exhaust manifold or turbocharger. Inside the ATS cabinet, any burnt smell that was absent before the test is a work order item—it may indicate contact arcing on a main power contactor. Also check the floor beneath both the generator and the ATS for any new fluid that appeared during the run. Photograph all findings before packing up your tools.

ContextA single out-of-spec reading may be an instrument error; a consistent 10% deviation from the previous month's value in the same parameter is a trend that demands action. For battery voltages, a drop of more than 0.2 V in float voltage from last month on a specific unit is significant. For generator parameters, a 10% increase in transfer time from month to month warrants investigating timing relay calibration. Create a 'Flag' column in your log template—circle or highlight any value that triggers this threshold so the reviewing supervisor can identify it at a glance without reading every number in the table.

ContextThe remarks section is the most legally and operationally important part of the log. During a Joint Commission, OSHA, or insurance audit, inspectors review this section first. A test showing perfect results with an empty remarks field is immediately suspicious—real equipment always produces something worth noting. Even a perfect test warrants an entry: 'No fault codes observed. Battery terminals on Unit 3 showed surface oxidation—cleaned, retreated, and retorqued. All parameters within spec.' This demonstrates genuine engagement with the equipment. Any deferred deficiency must reference a specific work order number and a target completion date.

ContextA signed log has legal standing; an unsigned one does not. The supervising engineer's signature confirms independent review of the recorded data, not just the technician's self-sign-off. Both signatures must be obtained before leaving the room—backfilling signatures days later is an audit flag and, in some jurisdictions, may constitute falsification of maintenance records. If a supervisor is not available for an in-person signature, a documented digital signature through your CMMS or an email acknowledgment with a timestamp is the minimum acceptable alternative.

ContextThe on-site binder must be physically present in or immediately adjacent to the equipment room—inspectors will ask for it on-site during surprise inspections, not in a central office filing cabinet. The CMMS upload creates the digital chain of custody and enables month-over-month trend reports that would be impossible to generate from paper alone. If your facility uses a building automation system integrated with a records management platform, configure an automated alert if the monthly test log is not uploaded within 48 hours of the scheduled test date—this catches human error before it becomes a compliance gap during the next audit cycle.

ContextCreating a work order on-site, in the moment, has a dramatically higher follow-through rate than a note transferred from a paper pad later in the day. Use your mobile CMMS app or a facility radio to place the work order before leaving the room. Assign priority based on operational impact: Priority 1 (safety-critical, resolve within 24 hours) for swollen batteries, burning odors in the ATS cabinet, or total transfer times exceeding NFPA 110 limits; Priority 2 (operational impact, resolve within 7 days) for low electrolyte, minor terminal corrosion, or elevated fuel consumption; Priority 3 (monitoring, resolve within 30 days) for minor trends that haven't yet crossed specification thresholds. A work order without an assigned owner and a priority level is just a wish.