Emergency Exit & Egress Lighting Monthly Battery Duration & Illumination Output Test

ContextHold the fixture face and look for cracks in the lens cover, warped or discolored housings (a sign of heat exposure or UV degradation), and any visible water staining inside the cover. A cracked lens reduces effective light output by up to 30% before the lamp itself shows any sign of failure — the unit will pass a visual illumination check but fail a lux meter reading. In wet locations (stairwells, loading docks, parking garages), moisture in the housing can corrode contacts and accelerate battery failure. If you see condensation inside the lens, the unit must be replaced, not just dried out — the seals are compromised.

ContextThe steady green pilot light is the fixture's self-reported 'battery is charging and healthy' signal. A dark pilot lamp means one of three things: AC power to the unit is interrupted (check the circuit breaker), the charger circuit has failed, or the battery is so degraded that it can no longer accept a charge. A blinking green light on units with microprocessor-based charging typically indicates a battery problem code — consult the manufacturer's manual for the blink pattern meaning. Do not assume a dark pilot lamp is a minor issue; it means the battery is either not charging or absent. Log it as a failure immediately and do not count the unit as compliant until it is repaired and the pilot lamp returns to steady green.

ContextOn units where the housing opens without tools (most do with a standard key or coin slot), briefly open the battery compartment and inspect the terminals. White or greenish powder on the terminals is sulfation or electrolyte leakage — both indicate the battery should be replaced. A loose connection on the terminal can cause the unit to pass a monthly functional test (the brief current draw seats the contact momentarily) but fail a sustained load test, because vibration under load opens the gap. If you see corrosion, photograph it, note the unit ID, and tag it for battery replacement within 7 days. Do not apply corrosion inhibitor and close — the underlying battery is already compromised.

ContextNFPA 101 requires emergency lighting to illuminate the egress path to a minimum average of 1 foot-candle (10.76 lux) at floor level, with a maximum-to-minimum ratio no greater than 40:1. A shelving unit, temporary partition, or stacked boxes within the beam spread of the fixture can create a dark pocket that violates both requirements even if the fixture itself is working perfectly. Check also that the fixture has not been painted over (common during renovation touch-ups) — a coat of paint on the lens can reduce output by 50–70% and is invisible to a quick visual check until you read the lux meter.

ContextMost modern fixtures have a recessed test button that simulates an AC power failure by opening the charging circuit. Press and hold for 3 seconds — this is the standard 'simulated failure' test per NFPA 101 Section 7.9.3. For older fixtures without a test button, interrupt AC at the breaker or use an in-line test switch if installed. Never test by physically disconnecting the battery — that bypasses the charging circuit logic and does not represent real-world failure conditions. Start your stopwatch the moment the fixture switches to battery power and note the exact time in the log.

ContextThe fixture must reach full output instantaneously upon AC loss — NFPA 101 allows no startup delay for emergency lighting (unlike emergency generator backup, which has a 10-second window). If any LED segments in an array fixture fail to illuminate, the unit fails, even if the remaining segments appear bright. Partial lamp failures are commonly caused by individual LED driver failures within multi-segment fixtures; the battery and charger may be fine, but the output is non-compliant. Note the exact failure mode in your log: 'two of six LED panels dark, lower right corner' gives the repair technician actionable information and creates a documented trail for warranty claims.

ContextWhile the unit is on battery, walk the full egress path segment that the fixture is intended to cover. Crouch or bend at the knee to simulate floor-level visibility — the human eye standing upright tends to perceive illumination as adequate even when floor-level foot-candles are below code. A dark spot at a floor-level transition (a step, a ramp, a door threshold) is the highest-risk condition because it's where trips and falls occur during evacuation. If you see a dark transition, note the location with reference to the nearest fixture ID so the spacing can be evaluated — the issue may be a missing fixture rather than a failed one.

ContextAfter releasing the test button or restoring AC, the pilot lamp should return to steady green within 30 seconds on most fixture types. A fixture that stays dark after AC restoration, or that shows a fault blink pattern, has a charger circuit failure. A unit that restores but then goes dark again within 10 minutes has a severely depleted battery that cannot sustain even the charging load — this is a unit that will fail during a real emergency even though it passed the initial functional test. Log the restoration behavior explicitly: 'restored to steady green within 15 seconds' or 'returned to blinking amber — charger fault suspected.'

ContextPlace the lux meter sensor flat on the floor (face up, parallel to the floor surface) at the midpoint between two fixtures along the egress path. NFPA 101 Section 7.9.2.1 requires a minimum average of 1 foot-candle (10.76 lux) at the floor. Take the reading while the fixture is on battery power, not on mains — charger-boosted output can exceed battery output by 15–25% in some fixture designs, which would give a falsely optimistic reading. If your meter reads in lux, divide by 10.76 to convert to foot-candles. Record both the lux value and the foot-candle equivalent in your log.

ContextThe midpoint reading tells you the average; the minimum reading tells you the worst case. NFPA 101 requires that the maximum-to-minimum illuminance ratio along the path does not exceed 40:1. If your brightest point (directly below a fixture) reads 10 foot-candles and your dimmest point reads 0.2 foot-candles, your ratio is 50:1 — non-compliant — even though both points individually exceed the 0.1 foot-candle minimum. Identify the dimmest point on each floor segment by walking the path with the lux meter running and noting the lowest reading. This is almost always at door thresholds or mid-corridor between widely-spaced fixtures.

ContextThese are the four locations where NFPA 101 requires individual measurement verification: changes in floor level, intersections of corridors, stairwells (each landing), and exit discharge areas. Evacuees slow down and navigate at these points — inadequate lighting here causes the highest injury risk. Stairwell measurements are taken at the center of each landing, not the top of the stairs. Exit discharge areas (the path from the exit door to a public way) must also be illuminated; many facilities forget that the obligation extends outside the building envelope to the point of final egress.

ContextNFPA 101 Section 7.9.3 mandates two separate test regimens: a functional test of at least 30 seconds every 30 days, and a full-duration test (90 continuous minutes) once every 12 months. These are not interchangeable. The 30-second test verifies that the fixture activates; the 90-minute test verifies that the battery has sufficient capacity to sustain code-required output for a full evacuation scenario. Your monthly log must clearly indicate which test type was performed for each unit. Marking every entry as a '90-min test' when only a functional test was performed is a falsification of records — a serious liability if reviewed by an AHJ or insurer after an incident.

ContextA full 90-minute battery drain test in an occupied building requires either: (a) testing one zone at a time with other zones providing ambient light, or (b) scheduling during nights, weekends, or building shutdowns. For the duration of the test, the fixture is running on battery and providing emergency-level illumination only — occupants navigating that zone during the test will experience reduced light levels. Brief your security team, post temporary signage at zone boundaries, and carry a portable task light. Buildings in healthcare or high-occupancy assembly occupancies should use approach (a) exclusively — never leave entire floors at emergency-only illumination levels during business hours.

ContextFor units with accessible battery terminal connections, use a voltmeter to record open-circuit or load voltage at each interval. A healthy sealed lead-acid battery in a fixture will typically start at 6.0–6.4V (for a 6V pack) and decline gradually; an unhealthy battery will drop sharply after 30–45 minutes, often below the fixture's low-voltage cutoff, causing it to shut off before 90 minutes. Logging voltage at intervals creates a degradation record: if a unit shows 5.8V at minute 30 this year versus 6.1V last year, you have early warning of approaching failure before the unit fails a full-duration test. Not all fixtures allow terminal access — if yours do not, at minimum log whether the unit passed (illuminating at full output at the 90-minute mark) or failed (lamp extinction or visible dimming before 90 minutes).

ContextMany modern fixtures include built-in self-test microcontrollers that automatically run a 30-second test at programmed intervals and record the result in onboard memory or transmit it to a central monitoring panel. These units are not exempt from manual verification — NFPA 101 still requires that a qualified person confirm the self-test log is current, that the last test result was a pass, and that the unit has not been in a fault state. Access the unit's self-test log (usually via a short press of the test button followed by a blink-code readout, or via the central monitoring software dashboard). If the last self-test was more than 32 days ago, the automated system has a scheduling fault — log this as a deficiency regardless of the pass/fail result.

ContextTest exit signs the same way as emergency luminaires: initiate battery mode and inspect every character and arrow for full, even illumination. A common failure mode in LED exit signs is the loss of one or two LED strings within a panel — the sign appears illuminated from a distance but on closer inspection shows darkened areas within the letters. The word 'EXIT' must be legible at the required viewing distance (100 feet per IBC Section 1013.6.1) under battery power. If any portion of the lettering is dim or dark, the sign fails — partial illumination of exit signage has been cited in multiple wrongful death cases where evacuees bypassed exits they couldn't clearly read during smoke-filled conditions.

ContextIn many older facilities, exit signs and egress luminaires share the same battery backup circuit. This creates a single-point failure risk: if the shared battery depletes before 90 minutes, both the path lighting and the exit sign go dark simultaneously. In newer installations, exit signs are on dedicated circuits with their own batteries. Check your electrical drawings to confirm the circuit topology, and if exit signs share a battery with luminaires, calculate whether the combined load current is still within the battery's 90-minute capacity rating. A battery rated for one luminaire at 3W may only sustain 60 minutes when also powering a 5W exit sign on the same circuit.

ContextStand 100 feet away from each exit sign (or the maximum unobstructed line of sight in the space, whichever is shorter) and verify the sign is clearly readable. Intervening obstacles — hanging banners, temporary partitions, merchandise displays in retail occupancies — can reduce effective viewing distance to below the code minimum. Also check that no new signage has been installed adjacent to an exit sign that could create confusion about the direction of egress. An exit sign pointing left next to a 'Fire Extinguisher →' sign pointing right has been documented as causing hesitation delay during drills.

ContextEvery entry must be individually traceable — 'all units on Floor 3 passed' is not acceptable log format for AHJ review. Each unit needs its own line with a unique identifier (asset tag, serial number, or grid reference) and a specific location description (e.g., '2nd floor east stairwell, top landing, north wall'). The tester's name and employee ID must be legible. If you use a paper log, print in block letters. If you use a digital system, ensure the entry is time-stamped by the system, not self-reported. An entry with no tester name is treated as unverifiable by AHJ inspectors and may not count toward compliance for that unit.

ContextA log that reads 'Pass' for illumination output is legally defensible only if your facility's internal standard defines what pass means in measurable terms and that standard meets or exceeds NFPA 101's 1 foot-candle minimum. A log that reads '1.3 fc at midpoint, 0.8 fc at door threshold — both pass' is far stronger: it documents the actual measurement, identifies where in the path the reading was taken, and demonstrates that a calibrated instrument was used. Over multiple months, numeric readings reveal degradation trends before the unit formally fails — a fixture that reads 1.4 fc in January, 1.1 fc in April, and 0.9 fc in July needs to be investigated before the September reading falls below 1.0 fc.

ContextUse a consistent, unambiguous marker system — a red 'F' in a circle works well on paper logs; a red status badge in digital systems. More important than the flag color is the failure description: 'battery dead — no illumination on test' tells the repair technician something different than 'illuminated but reading 0.4 fc — likely failed lamp' or 'unit activated but shut off after 45 minutes — battery duration failure.' Precise failure mode descriptions reduce diagnostic time and prevent technicians from replacing the wrong component. A battery replacement won't fix a lamp failure, and vice versa.

ContextA skipped unit is not a passed unit. If a fixture is inaccessible — the room is locked, scaffolding is in the way, a tenant has restricted access — enter it in the log as 'access unavailable — makeup test scheduled' with the specific date the makeup test will occur. Do not leave the field blank or omit the unit from the log. From a compliance standpoint, an untested unit is equivalent to a failed unit: you cannot certify the egress system as compliant while units remain unverified. For facilities with persistent access challenges (leased tenant spaces, server rooms requiring escorted entry), build a formal access protocol into your testing SOP and document it as standard practice.

ContextAttach a red 'OUT OF SERVICE — INSPECTION REQUIRED' tag to the fixture before leaving the area. This serves two purposes: it prevents a subsequent test cycle from missing the unit's failed status, and it alerts any AHJ inspector or insurance auditor who walks the space that the facility is aware of the deficiency and has documented it — proactive tagging typically reduces citation severity compared to an undocumented failure found by the inspector. The tag should include the date of failure, the failure mode, and the unit ID. Pre-printed tags cost less than $0.15 each; leaving a failed unit unlabeled in a busy facility risks it being counted as compliant by a different tester the following month.

ContextA single failed unit in a densely lit corridor may not create a code-non-compliant dark zone — but a failed unit in a low-fixture-density area (a stairwell landing with one luminaire, a remote corridor with widely spaced fixtures) can leave a section of the egress path below the 0.1 foot-candle floor minimum. Calculate whether the remaining fixtures can maintain coverage at that point; if not, a battery-operated temporary egress light (available for $30–$80 at electrical supply houses) must be installed in the same location within 24 hours. Do not wait for the repair work order to be scheduled — the egress path must be compliant while the permanent unit awaits service.

ContextYour facility's corrective action window — typically 30 days for non-emergency deficiencies under most AHJ frameworks — starts the day the failure is documented, not the day the work order is submitted. A failure logged on the 1st with a work order submitted on the 10th leaves you with only 20 days to complete the repair before the next monthly test cycle, during which the unit must be confirmed passing. Create the work order the same day and specify: unit ID, location, failure mode, whether a temporary unit has been installed, and the required completion date. Copy the facilities manager and the life safety officer so the work order cannot be quietly closed without confirmation of the fix.

ContextA repair is not confirmed until the unit passes a re-test under the same conditions as a standard monthly test: functional test, lux reading, and pilot lamp verification. Log the re-test as a separate entry and cross-reference both the original failure entry and the work order number. This linkage is critical during AHJ audits and insurance reviews — auditors specifically look for closed-loop documentation: failure detected → work order raised → repair completed → re-test passed. A repair that has no corresponding re-test entry leaves the failure technically open in the record, regardless of whether the unit is physically working.

ContextDivide the number of passing units by the total tested (exclude skipped units from the denominator — they are accounted for separately) and express as a percentage. A pass rate consistently above 95% with isolated random failures suggests normal end-of-life attrition. A rate that drops from 97% to 89% in a single month suggests a systemic issue: a shared battery supply batch failure, a circuit problem, or a widespread lamp type issue. Tracking pass rate as a trend line rather than a single monthly snapshot allows you to detect system-wide degradation months before individual unit failures begin to cascade. Target a pass rate of 98% or above as your internal facility standard.

ContextGroup your failed units by location and look for clusters. Three failures on Floor 6 out of 30 units (10% failure rate) versus one failure each on Floors 1–5 (2% failure rate) points to a Floor 6-specific issue: a shared circuit with voltage irregularities, a physical environment problem (higher humidity, more vibration from HVAC equipment), or a batch of batteries from the same installation date that are simultaneously aging out. Identifying clusters before running individual repairs prevents the inefficiency of replacing batteries one-by-one on a floor that will require a complete battery replacement program within six months anyway.

ContextPull the installation dates for each fixture from your asset register and calculate the average battery age. Sealed lead-acid batteries in egress fixtures have a rated service life of 3–5 years; nickel-cadmium packs typically 4–6 years. If more than 30% of your batteries are within 12 months of their rated end-of-life, plan a phased replacement program rather than waiting for individual failures. A proactive replacement program costs significantly less than the combination of emergency labor, temporary unit procurement, elevated AHJ violation risk, and the administrative overhead of managing dozens of open corrective actions simultaneously. Budget planning is also easier when battery replacement is scheduled rather than reactive.

ContextNFPA 101 Section 7.9.3 requires that records of tests and inspections be kept and made available for examination. Most AHJ offices expect to receive records within 24 hours of a written request. Digital logs stored in a shared drive accessible to the facilities manager, the safety officer, and a designated backup contact satisfy this requirement. Paper logs in a single binder at the facilities desk do not — if the facilities manager is unavailable or the binder is lost, you cannot produce the records in time. For multi-building portfolios, ensure a master log index exists that maps each building's records to a retrievable location. Losing records after an incident is legally treated the same as having no records.