Commissioning Fire Alarm Systems
• Design review
• Test procedures
• System readiness checklists
• Fire alarm program review
• Documentation of testing
• O&M manual review
• Warranty period review
Commissioning provides a fully functional fire alarm system, documentation of the fire alarm testing and documentation of how the fire alarm system functions to the building owner and operators. The CxC writes test procedures for the fire alarm system based on the design documents. The test procedures provide instructions to verify that the fire alarm system generates the design specified responses to specific inputs. Inputs to the fire alarm system consist mainly of detection devices, but some inputs consist of fault conditions that provide warnings to the building operators.
If the commissioning plan includes participation of the CxC during design development, then the CxC encourages formal communication between the designers and the building owner. The formal communication includes an owner’s project requirement (OPR) document, which lists the specific expectations of the building owner for the fire alarm system.
During design review, the CxC reviews the specifications and drawings based on the expectations listed in the OPR. Also, the CxC usually reviews the design for clarity of the requirements and for adequacy of testing requirements. An adequate testing requirement would be: Test the fire alarm system in accordance with NFPA-72 (National Fire Alarm Code) and verify all the system reactions per the direction of the CxC.
The CxC usually does not review the design with respect to code-related requirements such as the location and number of fire alarm/detection devices. The fire marshal or other authorities having jurisdiction review the design for code-related requirements.
After the design phase, the CxC prepares a system readiness checklist (SRC) and a fire alarm test procedure. The SRC contains items that need to be marked off as complete by the fire alarm contractor prior to the start of testing. The fire alarm test procedure contains items that the CxC intends to witness.
The intent of the SRC is to document that the fire alarm system is complete. The intent of the test procedure is to verify that the system is complete. The obvious distinction between the SRC and the test procedure is that a biased party (the contractor) completes the SRC and an owner’s representative (CxC) verifies completion of the test procedure.
To a large extent, the CxC develops the SRC based upon past experiences of finding incomplete work during testing. Incomplete work wastes the time of all the parties involved with testing. Thus, the SRC contains such items as verifying:
• All devices are installed
• All devices are wired to the control panel(s)
• All devices’ electronic addresses are properly set
• All programming is complete
• All interfaces with the sprinkler system are complete
• No trouble indications are displayed at the control panel(s)
Ideally, the CxC prepares a fire alarm test procedure for issue with the construction bid documents. The test procedure would be a contract requirement that the bidders have had a chance to read and to develop a cost estimate for performance of the test. However, the time period between the completion of the design documents and the design issue for bid usually does not leave time to develop a test procedure. Thus, the CxC includes an example test procedure in the bid documents. The example test procedure contains all the elements of the required testing.
The CxC develops the final test procedure for use in the field towards the end of the construction project. While preparing the fire alarm test procedure, the CxC reviews miscellaneous project documents in addition to the bid documents’ requirements. The miscellaneous project documents include the fire alarm system submittal, design documents addenda, changeorders, and requests for information.
The CxC test procedure includes all test items indicated in NFPA-72 that are applicable to the specific installation plus the specific requirements of the project documents.
The CxC test procedure includes action statements for testing individual devices such as “activate all smoke detectors,” “activate all pull stations,” or “simulate a fault condition.”
Prior to testing, the CxC develops detailed spreadsheets that list all the devices to be activated and all fault conditions to be simulated. Each item of each list needs to be marked as complete by the CxC during testing.
In addition to documenting individual device tests, the test procedure needs to include a means to document fire alarm system responses to device activation. Often, the design documents indicate the required system reaction in a matrix (table) format.
The various categories of devices would be in the first column of the matrix. The various system reactions would be in the first row (or heading) of the matrix. The interior boxes of the matrix would be marked with an “X” to indicate which reactions are required for which device activation. The test procedure could include a blank version of the matrix to be marked-up during testing. Or, the test procedure could contain a list of each device category with the required system reactions itemized (sub-list) for each device category. Each sub-list item needs to be marked complete by the CxC during testing.
A CxC becomes an expert at developing lists. For large fire alarm systems with many devices, the CxC should request the fire alarm contractor to print out a list of devices in the system. This list can be used for mark-up during testing instead of spending hours developing a separate list in an original spreadsheet.
A typical fire alarm system test includes:
• Smoke detectors
• Strobe lights
• Horns and/or speakers
• Flow switches (sprinkler water)
• Tamper switches (sprinkler water valves)
• Control panels
A typical fire alarm system provides signals to provide the following for an alarm event:
• Audible and visual alarm
• Smoke damper closure
• HVAC shutdown
The start of fire alarm system testing usually involves the fire marshal (or other authority have jurisdiction) and the CxC. The fire marshal’s testing has priority, and he will direct the testing. However, many fire marshals perform limited testing. The CxC documents the testing performed by the fire marshal, and then after the fire marshal completes his testing, the CxC directs the remainder of the testing.
The following paragraphs describe testing as if the CxC directed the entire testing and in roughly the sequence that the author of this article would conduct testing.
After a test coordination discussion with the fire alarm contractor, the CxC requests the activation of a detection device such as a smoke detector or pull-station. Smoke detectors are activated using “canned smoke” (a special spray) and, obviously, the pull station is activated by hand.
After the device activates, the CxC walks the entire building to verify that all the strobe lights and horns (or speakers) function. The CxC carries the fire alarm layout drawing in order to verify the devices’ locations are per the design. The CxC uses a decibel meter to measure alarm sound levels in building spaces that seem relatively quiet. Fifteen decibels above ambient noise levels is the normal acceptance criterion. Often, most building areas have obviously adequate alarm sound levels so that decibel measurements are not necessary in all areas.
After verification that the strobe lights and horns function the fire alarm system response can be “silenced” (an override at the control panel) for testing all the remaining detection devices. This testing will verify for each device that it does activate and provides the proper signal to the fire alarm control panel: device electronic address and device name. The device electronic address is a unique identification number and the device name is a brief description of the device. The device name usually includes the device location (room number) and the device type (smoke detector or waterflow switch).
The CxC can verify that each detection device will initiate an alarm event by reviewing the fire alarm system program. For a simple fire alarm system, the program contains a list of each detection device (listed by device address). The fire alarm program provides the same system response (alarm) for the activation of each device in the list. The CxC verifies that the list contains each detection device shown on the fire alarm layout drawing.
Depending upon the availability of the fire sprinkler system contractor, the CxC will test the flow and tamper switches before or after testing all the other detection devices.
For flow-switch testing, the sprinkler contractor opens the flow-test valve. The sprinkler water flows through the sprinkler system pipe into a drain pipe (which is often permanently installed for sprinkler flow testing). The alarm signal to the fire alarm system from the flow switch must occur in less than 60 sec and preferably after a delay of about 25 sec or more. A flow switch adjusted to alarm quickly (5 to 10 sec) could cause spurious future alarms during normal building use because in some sprinkler pipe systems’ water pressure surges cause enough water flow to activate the flow switch. The CxC tests tamper switches by opening and closing the valve with the tamper switch.
The acceptance criterion is that the tamper switch must activate prior to the valve wheel being fully rotated three times from the full-open position. Tamper switch activation initiates a supervisory condition at the fire alarm control panel to indicate that the valve is closed (or more precisely, not full-open). The CxC should fully close the valve in order to verify that the tamper switch remains activated through the entire motion of the valve steam. Tamper switches are not high-tech devices, and sometimes the installation configuration causes the tamper switch to indicate the valve is closed at “three rotations” but open for the full-closed position of the valve stem.
If the design requires the fire alarm system to provide shutdown signals to smoke dampers and HVAC equipment (AHUs), the CxC needs to verify these system reactions. For some systems, any alarm device activation causes all smoke dampers and HVAC equipment to shutdown. For other systems, only duct smoke detectors will cause smoke dampers and HVAC equipment to shutdown. If any device causes shutdown, then the CxC activates one device and checks all the smoke dampers for closure and HVAC equipment for shutdown.
As similarly explained for the strobe and horn activation, the fire alarm program contains a list of alarm devices that cause smoke damper and HVAC equipment shutdown. The CxC reviews the program list to verify that list contains all the alarm devices (addresses).
Duct smoke detectors are important specialty devices to detect smoke in the HVAC system. Normally, the duct smoke detector consists of a smoke detector (head) in a plastic box mounted on the duct with two air probe tubes (metal) connected from the plastic box to the interior of the duct. One air probe tube traverses the width of the duct and has air holes that face into the airflow (and the end of the tube plugged). The other air probe tube does not have any air holes except that the end of the tube is open. The path of airflow (and any smoke) is into the holes facing the airflow to the plastic box and back into the duct via the tube with the open end. Smoke passing through the box activates the smoke detector.
Testing them consists of three parts. The CxC tests the smoke detector by spraying canned smoke onto the detector (just like normal smoke detectors) and then checking for smoke damper and HVAC shutdown. Then, after system reset, differential air pressure between the two tubes is measured. This involves opening the plastic box and attaching the air pressure gage plastic tubes to the open ends of the air probe tubes which enter the box. Attachment of the plastic tubes requires adaptors made out of a plastic (or cork) plug with a hole and metal fitting for the plastic tube. The required minimum differential pressure is 0.015 in. wc for most duct smoke detector configurations. A typical maximum differential pressure is 2 in. wc, which essentially is never an issue.
The third part of the duct smoke detector testing pertains to review of how the airflow is controlled in the duct. Many HVAC systems have variable airflow. The CxC needs to determine the HVAC operating mode with the lowest airflow, and CxC activates that HVAC operating mode, repeating the differential airflow measurement to verify the differential pressure is above minimum. It is not unusual for duct smoke detectors configurations to fail this low-airflow test.
With the field devices and fire alarm system responses tested, the CxC verifies field wiring fault detection features of the fire alarm control panel(s). The control panel will provide “trouble” signals to the building operators if a field wiring circuit has: (a) an open circuit, (b) a grounded circuit, (c) a short circuit, or (d) a smoke detector head that is not installed on the detector base.
The CxC simulates an open circuit by disconnecting one wire at the control panel terminal block for the field wiring circuit; grounded circuit by connecting (“jumping”) a wire from one field circuit wire at the control panel terminal block to the (grounded) control panel enclosure; and short circuit by connecting (jumping) both ends of a field wiring circuit at the terminal block. Creating a short has a slight potential of damaging control panel interior solid state devices, and permission should be sought from the fire alarm system technician beforehand. The CxC tests for a missing detector head by removing the smoke detector head from its base. Experience has indicated that these fault detection features essentially never fail to properly provide trouble signals.
Complex fire alarm systems in large buildings consist of some or all of these attributes:
• Hundreds of smoke detectors
• Zoned strobes and horns
• Zoned HVAC shutdown
• Zoned smoke damper closure
• Door closure (magnetic devices)
• Elevator recall
• Smoke control
For large systems with hundreds of detectors and many activated devices (dampers, doors, HVAC, smoke control) it is impractical to physically observe all system reactions for each of the many detectors. If a system had 100 detectors, and each detector caused (on average) 10 smoke dampers to close, five doors to close, and two HVAC units to shutdown, then physical observation of each device (damper, door, HVAC) for each detector activation would mean 1,700 observations. The following explains a relatively efficient method for testing.
A CxC will observe the action of each device within the complex system. At least once, the CxC will observe smoke damper closure, door closures, and elevator recall. Likewise, the CxC observes HVAC shutdown and smoke control startup in order to verify that control hardware components (wires and relays) create the desired action. In essence, the CxC observes that each electrical relay operated by the fire alarm system does initiate the desired mechanical reaction. Once the mechanical reactions have been observed, then the operation of the fire alarm program needs to be verified.
For a large, complex fire alarm system, the fire alarm contractor writes a fire alarm control program to operate the system in accordance with the design requirements. As mentioned previously, the design documents often indicate the required system reaction in a matrix (table) format. The design requirements indicate what system devices are activated based on which detector signals a fire event. Often, the detectors are grouped into zones and the fire alarm system devices are activated based upon which zone has a fire event. The fire alarm control program provides the logic for activating the devices.
For example, a high-rise building with smoke control could have each floor as a separate fire zone. The fire alarm control program operates smoke control equipment in a particular manner to contain smoke based upon which zone (floor) has a fire event. Activation of audible and visual devices by zones, instead of the whole building, is another example of a fire alarm control program function. In short, the fire alarm control program receives an input of a fire alarm and then generates signals to operate devices in the desired manner. All of the operated devices can be thought of as electrical relays that are either open or closed.
The link between the fire alarm control program and the physical devices (relays) of the fire alarm system is the addresses (identification numbers) of the devices. In a fire alarm system, each device has a unique address. The detectors have addresses and these are the inputs (on/off) to the program. The relays have addresses and these are the outputs (open/close) of the program. The fire alarm system will not respond correctly if the addresses are incorrect. During field testing, the CxC verifies the address of each device by observing that the address is correctly displayed on the fire alarm control panel when that device is activated.
For fire alarm program review, the CxC obtains a printout of the fire alarm control program. The program logic for each design-required fire alarm system reaction must be verified. For some programs this may mean tracing the program logic for each fire detector. Usually, for zoned fire alarm systems, tracing the program logic for each group (zone) of detectors verifies the control program function. For any program logic that is not decipherable, the CxC requests assistance from the fire alarm system technician (or programmer).
Many fire alarm systems have a history log that will show the time at which the system activated relays. The log can be reviewed after a simulated (or real) fire event to determine which devices (strobes, horns, dampers, HVAC, etc.) were activated. Obviously, this can be a great assistance to testing the system. Depending upon the size and complexity of the fire alarm system, reviewing the history log for the activation of each and every detector may be as efficient or more efficient than reviewing the fire alarm control program. Or, the history log can be used in conjunction with a listing of the fire alarm control program to analyze the fire alarm system.
Upon completion of field testing, the CxC reviews the testing documents for completion to ensure each device and system reaction has been checked off. For any equipment failures, the CxC provides a report to the project team.
The familiarity with the fire alarm system that the CxC obtains allows the CxC to perform a rigorous review of the O&M manual for the system. Also, if the commissioning plan includes training, the CxC can assist with the on-site training of the building operators.
During the normal one-year warranty period, the CxC periodically contacts the building operators to ask about system problems. If the system has persistent problems, then the CxC can assist with initiating and conducting a meeting with the contractors to address the problems. ES