Study guide

12.3 Reacceptance Testing After Hardware and Software Modifications

NICET Fire Alarm Systems Level III Exam

Track your progress as you read

Finished sections get a check mark.

1 Chapter 1: Certification Overview & Code Framework

1.1 NICET Level III Exam Structure, CBT Format & Open-Book Strategy 1.2 Candidate Prerequisites, Work History & Performance Measures 1.3 Primary Codes, Standards Hierarchy & AHJ Relationships

2 Chapter 2: Occupancy Classification & IBC Code Triggers

2.1 IBC Occupancy Classifications, Occupant Loads & Mixed Occupancies 2.2 Manual Fire Alarm System Triggers & Sprinkler Exceptions 2.3 Automatic Smoke Detection & Smoke Alarm Triggers

3 Chapter 3: Initiating Device Layout & Spacing

3.1 Smoke Detector Spacing, Ceiling Geometry & Beam Pockets 3.2 Heat Detector Spacing, High Ceilings & RTI Derating 3.3 Duct Smoke Detection & HVAC System Controls 3.4 Specialty Detection Technologies (ASD, Flame, Beam & Linear Heat)

4 Chapter 4: Notification Appliance Design: Audibility & Voice

4.1 Public vs. Private Mode Audibility & Ambient Sound Levels 4.2 Sleeping-Area 520 Hz Low-Frequency Signaling & Pillow dBA 4.3 Emergency Voice/Alarm Communications & STIPA Intelligibility

5 Chapter 5: Notification Appliance Design: Visual Notification

5.1 Wall-Mounted Visible Appliances & The Room Size Method 5.2 Ceiling-Mounted Appliances & Corridor Strobe Spacing 5.3 Strobe Flash Rate, Synchronization & Field-of-View Rules

6 Chapter 6: Electrical Circuits, Cabling & Pathway Survivability

6.1 NEC Article 760 Wiring Methods, Cable Types & Separation Rules 6.2 NFPA 72 Circuit Classes (A, B, C, D, E, N, X) & Fault Monitoring 6.3 Pathway Survivability Levels (0–3) & High-Rise Riser Protection

7 Chapter 7: Power Supplies, Battery Sizing & Voltage Drop

7.1 Secondary Power Battery Calculations & Standby/Alarm Sizing 7.2 NAC Voltage-Drop Calculations & End-of-Line Verification 7.3 Primary Power Branch Circuits, NAC Boosters & Battery Maintenance

8 Chapter 8: Building System Interfaces & Special Suppression

8.1 Elevator Emergency Recall Operation & Shunt Trip Interfaces 8.2 HVAC Duct Smoke Shutdown & Dedicated Smoke Control Systems 8.3 Sprinkler Waterflow, Valve Supervision & Clean Agent Interfaces

9 Chapter 9: Emergency Communications & Mass Notification

9.1 Mass Notification System (MNS) Design & Risk Analysis 9.2 Emergency Communications Signal Priority & System Override 9.3 Emergency Responder Radio Enhancement Systems (ERCES / BDA)

10 Chapter 10: Submittal Preparation, Shop Drawings & Plan Layout

10.1 Fire Alarm Shop Drawing Standards, Legends & Symbols 10.2 Submittal Package Assembly, Cut Sheets & AHJ Coordination 10.3 Construction Specification Review, Scope of Work & Site Evaluation

11 Chapter 11: System Commissioning, Acceptance Testing & Closeout

11.1 Installation Strategy & Sequence of Operations Matrix Testing 11.2 AHJ Acceptance Testing Protocols & Pre-Testing Procedures 11.3 As-Built Drawings, System Record of Completion & Handover

12 Chapter 12: Inspection, Testing, Maintenance & Field Management

12.1 Periodic ITM Frequencies, Methods & Smoke Sensitivity Testing 12.2 System Troubleshooting, Ground Fault Isolation & Impairment Procedures 12.3 Reacceptance Testing After Hardware and Software Modifications 12.4 Unwanted Alarm Management, Detector Siting & Verification Features 12.5 ITM Records, Retention Periods, Deficiency Reporting & Owner Responsibilities 12.6 Project Management, Crew Leadership & Subcontractor Coordination 12.7 Workplace Safety (OSHA 1926), Environmental Compliance & NICET Code of Ethics

12.3 Reacceptance Testing After Hardware and Software Modifications

Key Takeaways Explain This Topic AI

NFPA 72 (2022) Section 14.4.2 requires that modified fire alarm equipment be tested as though it had never been tested, together with all components, circuits, and system operations known to be affected by the change.

A site-specific software change requires testing 100 percent of the functions known to be affected plus an additional 10 percent of initiating devices not directly affected, not to exceed 50 devices.

The 50-device cap limits only the supplemental sample of unaffected devices; affected functions, circuits, and interfaces are tested without a cap.

The 2022 edition requires a qualified representative to be on-site at the protected premises to perform reacceptance testing following a remotely programmed site-specific software revision.

Reacceptance is not complete until the Record of Completion is revised and the as-built drawings, sequence of operations, and archived software match the installed system.

Last updated: September 2026

12.3 Reacceptance Testing After Hardware and Software Modifications

Core Overview: Acceptance testing happens once. Reacceptance testing happens every time anyone touches the system afterward — and it is where Level III technologists carry direct professional liability. NFPA 72 (2022) Section 14.4.2 defines exactly how much of a modified system must be retested, and the answer is never "just the device I replaced." This section covers the hardware rule, the site-specific software rule, the 10 percent sampling allowance and its 50-device cap, the 2022 remote-programming requirement, and the Record of Completion revision that closes the loop.

1. Why Reacceptance Exists

A fire alarm system is a single integrated machine, not a collection of independent devices. Swapping one addressable smoke detector changes the loop's electrical loading. Editing one line of cause-and-effect logic can silently strand an elevator recall output. NFPA 72 therefore treats modified equipment as never having been tested , and requires the modifying contractor to prove the system still performs its approved sequence of operations.

For the NICET Level III candidate this maps directly onto two published work elements — 3.1.3 oversee system commissioning and 3.2.2 resolve impairments/deficiencies — and onto performance measure 0303-5108 , which requires the candidate to plan and oversee a complete and successful system commissioning including documentation of test completion. +-------------------------------------------------------------------------+

  • | ACCEPTANCE vs. REACCEPTANCE |
  • | |
  • | ACCEPTANCE (new system) ===> 100% of devices, 100% of functions |
  • | No sampling. No exceptions. |
  • | |
  • | REACCEPTANCE (modified) ===> Everything changed + everything the |
  • | change could affect + a defined sample |
  • +-------------------------------------------------------------------------+
  • 2. The Hardware Rule (NFPA 72 14.4.2)

When initiating devices, notification appliances, or control equipment are added, replaced, relocated, or removed , the modified equipment is tested as new equipment, and so is everything the change could have affected.

What changed What must be tested

Initiating devices added, replaced, or relocated Each added, replaced, or relocated device, plus the control-unit functions it drives

Notification appliances added or changed The appliances themselves, plus circuit voltage at the new end-of-line and the affected NAC's load calculation

Control unit, power supply, or booster replaced All functions of the replaced equipment, plus the circuits it supervises

Devices removed Verification that the removal did not create an open, an unsupervised branch, or an orphaned output mapping

Anything on an affected circuit 100 percent of components, circuits, and system operations known to be affected

Important

Adding appliances is an electrical change, not just a device change. Hanging two more horn/strobes on an existing NAC changes total load, end-of-line voltage, and battery standby and alarm demand. Reacceptance for that change includes reworking the voltage-drop and secondary-power calculations, not merely watching the new horns sound.

3. The Site-Specific Software Rule and the 10 Percent Sample

Site-specific software is everything the installing technician programs into the panel for this building : device addresses and custom labels, zone assignments, input/output logic, timers, alarm verification settings, notification groups, elevator recall matrices, door release, fan shutdown, and off-premises routing.

When any of that changes, NFPA 72 requires:

100 percent of the functions known to be affected by the change shall be tested.

Plus an additional 10 percent of initiating devices that are not directly affected by the change, not to exceed 50 devices .

The Record of Completion shall be revised to reflect the change.

Three Misreadings That Fail Candidates

Misreading Reality

"Test 10 percent of the system and stop." The 10 percent is additional to the affected-function testing. It never substitutes for it.

"The 50-device cap limits the whole reacceptance." The cap limits only the additional sample of unaffected initiating devices. Affected functions, circuits, and interfaces are uncapped.

"Any sample of 10 percent will do." The sample should be selected randomly and should include at least one device per initiating device circuit.

Worked Example: Rezoning a Six-Story Office

A technologist edits the cause-and-effect matrix so that floors 4, 5, and 6 now sound on a common alarm group instead of individually, and adds a duct detector on the fifth floor to shut down AHU-3. The building has 420 initiating devices total. STEP 1: Functions known to be affected -> test 100%

  • New duct detector: alarm response + AHU-3 shutdown
  • Floors 4, 5, 6 notification grouping (all appliances on those floors)
  • Any annunciator text or off-premises point tied to those zones
  • STEP 2: Additional sample of UNAFFECTED initiating devices
  • 10% of 420 = 42 devices
  • 42 < 50 -> the cap does not bind; test 42 devices
  • Select randomly, at least one per initiating device circuit

STEP 3: Revise the Record of Completion and the as-built drawings

Had the building held 900 initiating devices, 10 percent would be 90 — but the cap limits the additional sample to 50 . The affected-function testing in Step 1 is unchanged and uncapped.

4. The 2022 Remote Programming Requirement

Off-site programming of control equipment is now routine. The 2022 edition of NFPA 72 responded by clarifying that where programming is performed remotely, a qualified representative must be on-site at the protected premises to perform the reacceptance testing that follows the site-specific software revision. A technician logging in from the office cannot self-certify a remote software change; somebody has to stand in front of the panel and prove the affected functions still operate.

5. Sequencing Reacceptance on a Live Building

Reacceptance is performed on an occupied, protected building, which changes the choreography completely compared with new construction.

Notify before you touch anything. The supervising station, the building owner or designated representative, and the AHJ where required. A reacceptance test that trips a central-station dispatch is a preventable and expensive error.

Place the affected portion on impairment and follow the impairment procedures — tagging, coordinator notification, and fire watch where the impairment exceeds the code threshold.

Pre-test your own work before anyone witnesses it. Field-verify the changed devices and functions, then run the affected-function list end to end.

Restore and confirm. Return the system to normal, verify no trouble or supervisory conditions remain, and confirm signal receipt with the supervising station before releasing the impairment.

Document. Revise the Record of Completion, update the as-builts and the sequence of operations matrix, archive the revised site-specific software to non-volatile media, and issue the test record.

6. Realistic Exam Traps

"Only the new device needs testing." The most common wrong answer on the exam. The code scope is the modified equipment plus everything the modification could affect .

Conflating the three different 10 percent rules. Site-specific software reacceptance sampling, executive (manufacturer) software changes, and the circuit-integrity test method in the testing table each have their own 10 percent framework. They are not interchangeable, and exam distractors deliberately swap them.

Forgetting the Record of Completion revision. Reacceptance is not complete when the devices test good; it is complete when the documentation matches the installed system.

Treating a battery swap as paperwork-free. Replacing a battery set changes secondary power capacity. The replacement is tested, dated, and recorded.

Skipping the supervising station call. Unannounced testing generates a false dispatch, which in many jurisdictions is a fineable offense against the building owner your company works for.