Executive Summary

Every generator no-start in a healthcare facility, data center, or mission-critical site starts the same investigation: pull the run log, check the last test, find out when the battery was replaced. In most cases the answer is the same. The battery read fine on the last monthly test. It still failed to crank the engine when the utility dropped. A 12.6-volt static reading is not a capacity reading, and NFPA has never treated it as one.

NFPA 110 Chapter 8.3 requires weekly inspection of batteries and monthly testing and recording of specific gravity for lead-acid batteries — with conductance testing permitted as a substitute. NFPA 111 extends parallel requirements to stationary UPS battery banks. This issue breaks down what the standards actually say, what surveyors look for in the testing log, and why a generation of facilities maintaining batteries by voltage alone are running a compliance and reliability risk they have not been pricing correctly.

01
The Code

What NFPA 110 & 111
Actually Require

Facilities managers often quote NFPA 110 from memory as "monthly test the generator." That misses the other half of Chapter 8 — the battery maintenance language is separate, specific, and strict. The requirements are minimum standards, not best-practice suggestions.

8.3
NFPA 110 — Chapter 8.3
Routine Maintenance and Operational Testing · Storage Batteries
Enforceable
Inspection
Weekly
Specific Gravity
Monthly
Defective Battery
Replace Immediately
Weekly Inspection
Storage batteries used in connection with EPSS equipment shall be inspected weekly and maintained in full compliance with manufacturer specifications. The weekly inspection includes electrolyte level (for vented/flooded cells) or battery voltage (for VRLA/sealed).
Monthly Testing
Maintenance of lead-acid batteries shall include monthly testing and recording of electrolyte specific gravity. Battery conductance testing shall be permitted in lieu of the testing of specific gravity when applicable or warranted — which is the path most facilities now use, especially for VRLA/sealed batteries where specific gravity cannot be measured.
Defective Batteries
Defective batteries shall be replaced immediately upon discovery of defects. There is no "we'll get to it on the next PM" provision. A defective battery in an EPSS is an immediate-action item under the code.
The Critical Word
"RECORDING" — testing alone is not compliance. Monthly test results must be recorded and retained. Surveyors will ask for the log. If you tested but did not document, from the code's perspective you did not test.

How NFPA 111 Extends the Framework

NFPA 111 covers Stationary Emergency Power Supply Systems (SEPSS) — the stored-energy world. That means UPS battery banks, central battery systems, and large stationary storage, not just generator cranking batteries. The testing framework runs in parallel to NFPA 110 but adds language appropriate to larger installations: battery room requirements, capacity testing protocols, and reference back to industry standards like IEEE 450 (for vented lead-acid) and IEEE 1188 (for VRLA).

Both standards are overseen by the same EPS-AAA Technical Committee, and the two move together. If you run a hospital with a diesel generator and a central UPS system feeding life-safety loads, you are subject to both standards at once. Surveyors will expect a single, coherent battery maintenance program that addresses both.

The practical rule: if a battery powers anything on the emergency branch — engine start, control circuits, generator excitation, UPS, central lighting — it falls under one of these standards. "We don't really track that battery" is not an answer a surveyor accepts.

02
Why Voltage Fails

Why a Voltage Check Is Not
a Battery Test

Walk into a hundred facilities and you will find a hundred battery logs that show a neat column of voltage readings — 12.6, 12.6, 12.7, 12.6 — every month for years. Then the generator fails to start during a real outage. The battery voltage was fine. The battery capacity was not.

Static voltage tells you the surface charge of the battery. It does not tell you how much current the battery can deliver under load, which is the only thing that matters when the starter motor pulls 400–800 amps to crank a diesel engine. A battery at the end of its service life, with heavily sulfated plates and reduced active material, can read 12.6V and still collapse to 8V the instant a starter engages.

!
What Actually Measures Battery Health
Test Methods Recognized Under NFPA & Industry Practice
Reference
Specific Gravity (Flooded Lead-Acid Only)
Hydrometer measurement of electrolyte density per cell. Detects sulfation, stratification, and low-charge conditions. Cannot be used on VRLA/sealed batteries. Required monthly for vented lead-acid under NFPA 110.
Conductance / Impedance Testing
AC conductance (or ohmic impedance) measurement that correlates to plate condition and capacity. Takes seconds per battery, non-invasive, trendable over time. Permitted in lieu of specific gravity under NFPA 110. Required practice for VRLA per IEEE 1188. This is the method most modern facilities should be using.
Load Testing
Applies a known load to the battery while monitoring voltage drop. A healthy starting battery should hold above a minimum cranking voltage (typically > 9.6V on a 12V system at 0°F) under full starter load. Performed during the monthly generator start on a cold engine, or via dedicated carbon-pile load tester.
Capacity / Discharge Testing (UPS/SEPSS)
For UPS battery strings, full capacity discharge testing per IEEE 450 or IEEE 1188 verifies runtime at rated load. This is the gold-standard test for stationary storage and is typically performed annually or every 2–3 years on a rotating schedule. Required to defensibly claim runtime capacity.

What surveyors look for: a battery log that shows only voltage readings, month after month, with no specific gravity entries and no conductance data, is a documentation deficiency under NFPA 110 Chapter 8.3. The standard does not say "test voltage monthly." It says test specific gravity or conductance monthly. Recording voltage in place of either does not satisfy the requirement.

03
Service Life

How Long Each Battery Type
Actually Lasts

The service-life numbers below are the range most engineering firms and battery manufacturers cite. Actual life in your facility depends on ambient temperature (every 10°C above 25°C roughly cuts VRLA life in half), charge voltage regulation, and how often the battery is cycled. These numbers are the windows inside which facilities should be planning replacement, not targets to ride to failure.

S
Starting Batteries (Lead-Acid)
Generator Cranking — Typical Service Life
3–5 Years
Typical Replacement Window
3 to 5 years is the honest window for standby generator starting batteries in North American climates. Commercial truck-grade SLI batteries deployed as generator starters trend shorter (3–4 years). Purpose-built standby starting batteries can reach 5 years with disciplined maintenance and charger regulation.
What Kills Them Early
Heat (primary factor). Undercharging from a marginal float charger. Overcharging that boils off electrolyte. Vibration that cracks plate welds over years. Deep discharges if the generator runs for extended cranking without the field starting.
Replacement Rule
Replace at 4 years unless conductance testing justifies extension. Do not ride a starting battery past 5 years. The cost of a set of batteries is trivial compared to the cost of a no-start during a real outage.
V
VRLA / AGM (Sealed Lead-Acid)
UPS Float Applications — Typical Service Life
3–5 Years Typical
Design Life vs. Service Life
Manufacturers sell VRLA batteries with nameplate "design life" ratings of 5, 10, or even 20 years. Design life is not service life. It is a laboratory rating at ideal conditions. Real-world service life in continuous float applications is typically 3 to 5 years for standard VRLA, 5 to 8 years for long-life product lines, and longer for specialty chemistries when actively managed.
Required Testing
Ohmic/conductance measurement semiannually at minimum (more often if trending upward). Visual inspection for bulging, leaking, corrosion. Capacity discharge testing every 2–3 years per IEEE 1188. Replace when conductance drops to 80–90% of reference value (manufacturer-specific threshold).
The Heat Problem
VRLA service life halves for every 10°C above the 25°C reference temperature. A battery room running 35°C in summer is burning through design life at double the expected rate. Battery room HVAC is a capital-planning line item, not a nice-to-have.
F
Flooded Lead-Acid (Stationary)
Large Vented Stationary Banks — Typical Service Life
10–20 Years
Typical Replacement Window
10 to 20 years with disciplined maintenance. Large vented flooded cells remain the longest-lived chemistry for stationary applications, which is why they are still specified in new substation, utility, and some data center installations despite the maintenance burden.
Maintenance Burden
Monthly specific gravity per cell, water top-off, terminal cleaning, equalization charging. Flooded banks require battery room ventilation per IFC and NFPA 1, spill containment, and dedicated PPE. The long service life is a trade-off for ongoing maintenance labor.
N
Nickel-Cadmium (Ni-Cd)
Industrial & Switchgear Applications
15–20 Years
Typical Replacement Window
15 to 20 years, sometimes 25+. Ni-Cd tolerates high temperatures and deep discharge better than lead-acid. Commonly found in switchgear tripping, utility substations, and harsh-environment applications. Higher upfront cost, lower lifecycle cost when runtime is amortized.
L
Lithium-Ion (LFP, NMC)
Newer UPS & EPSS Deployments
10–15 Years
Typical Service Life
10 to 15+ years in data center UPS applications. Lithium Iron Phosphate (LFP) chemistry is the preferred choice for standby UPS due to thermal stability. Significantly higher upfront cost than VRLA, but smaller footprint, better energy density, higher tolerance of partial-state-of-charge, and longer life.
Code Considerations
Lithium-ion ESS installations trigger additional code requirements under NFPA 855 (Standard for the Installation of Stationary Energy Storage Systems). Battery cabinet listings, thermal runaway mitigation, ventilation, and fire suppression requirements apply. This is not a drop-in VRLA replacement from a code standpoint.
Battery Failure Risk vs. Time in Service
Starting — 5+ Yrs
Critical
Starting — 4–5 Yrs
High
VRLA — 5+ Yrs
High
VRLA — 3–5 Yrs
Elevated
Starting — Under 3 Yrs
Standard
04
What To Do Now

Fixing the Battery Program
Before the Next Survey

Most of the battery-related findings we see in compliance gap reports come from the same three gaps: voltage-only logs, missing monthly specific-gravity or conductance entries, and batteries well past their service-life window with no documented replacement plan. None of these are hard to fix. They require a protocol change, a modest capital line, and honest record-keeping.

Action Items — Battery Compliance & Reliability
  • 01Audit your battery log today. Pull the last 12 months. Are there monthly specific-gravity or conductance entries, or only voltage? If the answer is "only voltage," your log does not satisfy NFPA 110 Chapter 8.3. Fix the log template before the next PM.
  • 02Standardize on conductance testing. For almost every modern facility, conductance is the right path: it works on both flooded and VRLA batteries, produces a trendable numerical reading, and is permitted under NFPA 110 in lieu of specific gravity. A battery conductance tester runs $400–$1,500 and pays for itself the first time it flags a battery before a failure.
  • 03Document install dates and replacement windows. Every starting battery and every UPS battery string needs a documented install date and a scheduled replacement date. For starting batteries, default to a 4-year replacement cycle unless conductance data justifies extension. For VRLA, replace at 3–5 years for standard product, with conductance trending as the decision input.
  • 04Do an annual load test on starting batteries. Specific gravity or conductance catches degradation early, but an actual load test — either via a carbon-pile tester or by observing cranking voltage during the monthly start on a cold engine — is the definitive capacity check. Once per year, during cold weather if possible. Record the results.
  • 05For UPS: schedule a capacity discharge test. Per IEEE 1188 for VRLA or IEEE 450 for vented, capacity discharge testing should be performed every 2–3 years minimum. If you cannot remember the last time your UPS strings were capacity-tested, schedule it. A "runtime rating" without a capacity test is marketing, not data.
  • 06Control battery room temperature. If your UPS room runs above 25°C average, you are burning through battery life at an accelerated rate. HVAC that holds 72–77°F is not optional for VRLA installations. Flag inadequate cooling as a capital-planning item with the battery replacement program.

Battery Compliance Readiness Checklist

Use this checklist to evaluate your facility's battery maintenance program against NFPA 110 Chapter 8.3 and NFPA 111 requirements before your next AHJ or accreditation survey.

  • Weekly Inspection
    Weekly battery inspection (electrolyte level or voltage, as applicable) is documented in the EPSS log with date and initials of the individual performing the inspection.
  • Monthly Testing
    Monthly specific-gravity readings (for vented lead-acid) or conductance readings (for VRLA or in lieu of specific gravity) are recorded. Voltage-only logs are a deficiency.
  • Install Date Documentation
    Every starting battery and UPS battery string has a documented install date on the battery itself and in the maintenance record.
  • Replacement Schedule
    A documented replacement schedule exists. Starting batteries on a 4-year default cycle. VRLA replaced at 3–5 years or when conductance drops below manufacturer threshold.
  • Annual Load Test
    Starting batteries receive an annual load test — carbon-pile or observed cranking voltage under actual start load. Results recorded.
  • UPS Capacity Test
    UPS battery strings have been capacity-discharge-tested within the past 2–3 years per IEEE 450 / IEEE 1188. Documentation on file.
  • Battery Room Environment
    Battery room temperature is controlled to 72–77°F average. Ventilation meets IFC/NFPA 1 requirements for the battery chemistry installed.
  • Defective Battery Response
    Written protocol exists for immediate replacement of defective batteries consistent with NFPA 110 requirement. No "wait until next PM" provision.
Uptime Compliance Services

Battery Program Gap Report

Our compliance gap report audits your entire battery maintenance program against NFPA 110 Chapter 8.3 and NFPA 111 — testing log review, conductance vs. voltage compliance, replacement schedule, and UPS capacity documentation. Find the gaps before the surveyor does.

Request a Gap Report Schedule Onsite Audit
The Compliance Brief
← Issue #005 All Issues Next Issue →