This month’s column describes the smoke management testing process during the construction of a 600,000-sq-ft prison. In brief, the prison’s smoke management system consists of: fire/smoke dampers, smoke barriers, life-safety fans (supply and exhaust), automatic controls, manual controls, and system status monitoring. The fire alarm system features manual control and status monitoring via a graphical interface system. The fire alarm system is the computerized processor for the smoke management system, receiving inputs and generating outputs.

Several members of the project team consider this to be the most complex smoke management system they have ever known.

A Change of Plan

The designers, smoke management specialist, and contractors reviewed detailed verification test procedures written by the commissioning consultant. The verification test procedures implicitly assumed that the smoke management system would be fully functional prior to testing. The verification test procedures represented an integrated method of system testing.

The contractors rejected the concept of integrated system testing with a fully functional system, piece-by-piece testing as sections of the system were completed. The project team proceeded with this testing method.

No Leniency Allowed

The commissioning consultant used itemized lists (see the June 1999 column) to track which system segments had been tested.

The testing team was comprised of the electrical contractor, fire alarm subcontractor, general contractor, and commissioning consultant. The team began the segmented testing by showing that all the initiating devices (smoke detectors and sprinkler water flow switches) generated the proper signal (device address) to the fire alarm control panels. Activation of each initiation device (perhaps 1,500) was a time-consuming process, lasting twenty days. At the end of this testing, all of the inputs to the fire alarm system were known to be correct.

Next, the team observed the closure of each fire/smoke damper, which closes upon loss of current to the fire/smoke damper actuator. Thus, for each fire/smoke damper, the wires to the actuator were disconnected in order to observe closure.

Testing of smoke barriers involved the use of specialized pressurization fans and differential pressure gauges. The contractors leased the equipment for use at the construction site. For each smoke zone, the contractors mounted the fans in the entrance to a smoke zone and pressurized the zone. In effect, the smoke barriers for each zone were considered acceptable if the smoke zone maintained adequate pressure. The team tested the operation of life-safety fans by generating an output (start) signal from the fire alarm system. The lifesafety fans were considered acceptable if the fans provided adequate air flow or provided adequate space pressurization.

With the testing completed for initiation devices, fire/smoke dampers, and life-safety fans, the team proceeded with tests of the fire alarm computerized processor (i.e., the control software). Testing consisted of generating a fire alarm (activation of an initiation device) in each smoke zone and observing the results: fire/smoke damper closure, alarms sounded, and life-safety fans operated. The (designer) specified results were different for each smoke zone. The control software was considered acceptable if the specified results were observed. Test team members “walked the building” in order to observe the results. Unfortunately, some results were incorrect and subsequent retesting needed to be conducted.

The final aspect of testing included manual control of life-safety fans and supervision of equipment status via the graphical interface system. Due to the overall complexity of the smoke management system, testing of manual control and status supervision was difficult and confusing. The team contacted the design engineer for clarifications with respect to the design intent. After the design consultation, testing resumed and proceeded normally.

An Exercise in Exercise

In retrospect, the segmented testing process was inefficient with respect to labor-hours. The testing team repeatedly visited areas of the facility as parts of the smoke management system became functional. If the graphical interface system had been functional at the start of testing, then the equipment status could have been checked without “walking the building.” However, the segmented testing process did allow an early start and contributed to a relatively early end to the testing process. ES

(Note: ASHRAE Guideline 5, “Commissioning of Smoke Management Systems” provides a full outline of the commissioning process.)