ASHRAE Journal - December 2011 - 18

A total of 20 UFAD tests were conducted according to the test Field Testing methods described previously: 18 Successfully executing the UFAD tests were Type I on 84 plenums leakage test and the described test covering 735,000 ft2 (68 284 m2), procedure involved extensive ﬁeld and two tests were Type II on ﬂoors quality control during installation, covering 102,000 ft2 (9476 m2). The cleaning, and preparation of the ﬂoor average leakage rate of the Type I before testing. During construction of Power and telecom conduits had their ends sealed. and Type II results are presented in underﬂoor production a full-time ﬁeld Table 1. person was assigned to ensure that the plenums were being built For Type I tests the average leakage for all Type I tests for to minimize leakage. This was done by ensuring that the bottom the entire 735,000 ft2 (68 284 m2) tested was 4.1%. A typical and edges of walls were sealed with plenum-rated caulking. Pen- test result for one of the levels tested is presented in Table 2. etrations such as power and telecom conduits that entered the The best result was 1.1% leakage and was achieved twice on plenum boundaries had their below ﬂoor ends’ sealed, as shown 15,950 ft2 (1482 m2) plenums. at right. These qualitative visual inspections proved vital in mainFor the Type II tests the average leakage for the 102,000 ft2 taining positive testing results, especially since troubleshooting (9476 m2) tested was at 7.6% of allowable. The same plenums was complicated by USACE requirements forbidding the use of are shown in Table 2 (Type I testing) and Table 3, (Type II smoke to determine leakage points. testing). Type II testing was conducted with all furniture and After construction of the pedestal support system and before spline wall penetrations, making this test a ﬁnal as-built result. the ﬂooring was installed, the underﬂoor was thoroughly vacu- The results show that ﬂoor leakage was 6.2% higher than the umed and mopped. Any necessary repairs were performed and Type I test on the same ﬂoor. a quality control close-in inspection was performed with the The difference in results between the Type I and Type II tests is owner’s representative. an average 3.3% increase in leakage above Type I leakage. This The concrete ﬂoor tiles were installed, and the entire surface was comfortably below the 10% allowable leakage rate for Type was thoroughly cleaned then covered in self-adhering carpet II tests. The other result for Type II leakage was a 1.3% increase masking plastic in 3 ft (0.92 m) wide strips. At the edges of a above Type I and was achieved on a 16,100 ft2 (1496 m2) plenum. plenum area and against walls a 2 in. (50 mm) wide packing tape was applied to seal the perimeter joint. Even with the addi- Lessons Learned tional care paid to these areas, the edge seals were a signiﬁcant During testing the project team learned how stringent the 5% leakage point for Type I testing. Empirically, it was discovered leakage rate could be and the level of quality control vigilance that a plenum with a low perimeter-to-surface area ratio per- required during construction. The ﬁrst ﬂoor of Type I testing took formed better than those with a high perimeter-to-surface area several attempts with remediation work in between testing efforts ratio. to obtain a successful result. These remediation efforts taught the
18 ASHRAE Journal ashrae.org December 2011

and sealed and once all plenums were ready they were turned on simultaneously and initial measurements were taken in each plenum. If the pressures were above the test pressure of 0.05 in. w.g. (12.3 Pa) the ﬂow rates through each rig were adjusted. Pressures below 0.05 in. w.g. (12.3 Pa) would have been considered unacceptable as they did not meet design conditions and immediate remediation was performed by ﬁnding and sealing any remaining leaks until the plenum pressure met or exceeded the required test pressure. All plenums tested were able to pass the required test pressure. The process for adjusting the fan volumes was complicated by slight leakage between plenums, the number of plenums on a ﬂoor, and the number of meters used to take readings. Lowering one plenum’s volume would decrease pressure in adjacent plenums with a ripple effect to interconnected plenums. The most accurate results are achieved when the plenums were at nearly the same pressure, factoring in equipment tolerances and time delays between readings. However, with such small pressures and manual adjustments, it was not practical to get all plenums to precisely the same pressure values.

Type II testing was performed after Type I testing was completed and the carpet and swirl diffusers were installed. The preparation required using 1 ft (30.5 cm) wide sections of carpet mask to cover the swirl diffusers and linear diffusers. Setup of the duct leakage test fans and measuring meters were the same as Type I testing with the exception being the increase in maximum allowable leakage, which was 10% greater than that measured for the same area for Type I testing. Two Type II tests were conducted. The ﬁrst Type II testing level was a completed ﬂoor with carpet and swirl diffusers installed. The second Type II test was also on a completed ﬂoor with carpet and swirl diffusers, but this ﬂoor also had a furniture spline wall installed, and there were numerous throughﬂoor penetrations for power and telecom into the spline walls. The number of penetrations in the second Type II test required additional care in the pre-test prep work. A direct correlation was discovered between the time and caution spent with sealing and their related inspections and the resulting leakage rate.

Testing Results

ASHRAE Journal - December 2011

Table of Contents for the Digital Edition of ASHRAE Journal - December 2011