Comparative Analysis of Conventional and Underfloor Air

Document Sample
Comparative Analysis of Conventional and Underfloor Air Powered By Docstoc
					    Comparative Analysis of Conventional and Underfloor Air Distribution
    System Performance using the Air Diffusion Performance Index Method
                          (Extended Abstract of Technical Paper submitted for Publication)
                                                     James E. Woods 1 and Davor Novosel 2

      Expansion of air jets from supply air diffusers and grilles and corresponding reduction
of air velocity due to induction of room air is a fundamental principle of air diffusion that is
common for both conventional and underfloor air distribution (CAD and UFAD) systems.

           Based on this principle:
              • The primary objective of this analysis was to quantitatively evaluate the
                  efficacy of the ADPI method for use in UFAD systems by comparing the
                  results of ADPI data obtained in a test room configured as CAD and UFAD
                  systems under similar cooling loads and room reference point temperatures.
              • A second objective was to evaluate the practicality of using ASHRAE Standard
                  113-2005 for field-testing.
              • A third objective was to recommend modifications that will enable the use of
                  ADPI as a design tool for UFAD.

      To achieve these objectives, three replicate sets of air temperature and velocity data
were obtained for each of three cooling loads and for two system configurations in the
calibrated Building Technology Laboratory (BTLab) at the University of Nevada, Las
Vegas, as shown in Figures 1-3 (Tan et al 2008).

      The CAD and UFAD configurations were selected to maintain the same average air
temperature in the occupied zones (i.e., 75 +/- 1 °F) at three uniform cooling loads (i.e., 7.1,
15 and 30 Btu/h ft2) by varying the supply airflow rates and temperatures. The selection of
the type and locations of the four ceiling diffusers for the CAD configuration was based on
the design method described in ASHRAE 2005c to achieve an ADPI of at least 80% with a
uniform interior cooling load of 30 Btu/h ft2 in an interior area; evaluation for exterior loads
in perimeter areas was not included in this analysis.

      The 80% criterion was selected as it is consistent with the criteria for “acceptable
thermal environments” (ASHRAE 2004) and for “acceptable indoor air quality” (ASHRAE
2007). As a standardized design method of selecting the UFAD diffusers is not yet
available, eight swirl-type floor diffusers were selected, based on information from Bauman
(2003) and from manufacturers’ literature. Each diffuser was selected for a design airflow
rate of 110 cfm at the uniform cooling load of 30 Btu/h ft2. The return air grilles were
located in the ceiling at the four corners of the BTLab.
  Executive Director, The Building Diagnostics Research Institute, Inc., Chevy Chase, Maryland
  Chief Technology Officer, The National Center for Energy Management and Building Technologies,
Alexandria, VA
Extended Abstract                                               Page 1 of 10                8 August 2008
For use by NEMI
                                   Typical return air grilles 

                                   Typical lighting fixtures 

               Traversing mechanism 

                                   Typical Electric Resistance 
                                                                  Typical floor diffusers 

                             Figure 1 Test Room Configuration

Extended Abstract                      Page 2 of 10                          8 August 2008
For use by NEMI
     0          1               3                  5            7           9          11             13           15          17    19            21            23     25           27           29    30 ft
                                                                                                                                                                                                               0 ft
                                                       PE3: the third group of floor heaters, service area 200 ft2

                                                1               2           3          4           5           6                7        8         9           10     1 1            12
                                               A                                                                                                                                                               4

                                               B                                                                                                                                                               6


                                                       PE2: the second group of floor heaters, service area 200 ft


                                               E                                                                                                                                                               12



                                                       PE1: the first group of floor heaters, service area 200 ft2

                        Ceiling Diffuser                            Floor Swirl Diffuser                          Baseboard Heater                      T   Reference Point               Test Position

                                                                       Figure 2. Plan View of Test Grid in the BTLab

     0         1               3               5                7          9          11            13            15           17   19            21           23      25       27            29       30 ft
                                                                                                                                                                                      17 (104”)
          Zone C:                                                                                                                                                                     16 (98”)
          Upper zone in the test space
          6 to 9 ft above floor                                                                                                                                                       15 (92”)
          Induction zone in CAD systems                                                                                                                                               14 (86”)
          Stagnant zone in UFAD systems
                                                                                                                                                                                      13 (78”)
                                                                                                                                                                                      12 (72”)
                                                                                                                                                                                      11 (66”)
                      9 ft


                                                                                                                                                                                      10 (60”)

          Zone B:                                                                                                                                                                     9 (54”)
          Middle zone in the test space                                                                            T
          3 to 6 ft above floor                                                                                                                                                       8 (48”)
          Uniform mixed zone in CAD systems                                                                                                                                           7 (42”)
          Uniform mixed zone in UFAD systems
                                                                                                                                                                                      6 (36”)
          Zone A:
          Lower zone in the test space                                                                                                                                                5 (30”)
          0 to 3 ft above floor                                                                                                                                                       4 (24”)
          Uniform mixed zone in CAD systems                                                                                                                                           3 (18”)
          Mixing zone in UFAD systems
                                                                                                                                                                                      2 (12”)
                                               1               2           3           4          5           6                7    8             9         10        1 1       12    1 (4”)

                                                   Ceiling Diffuser                            Floor Swirl Diffuser                  T       Reference Point            Test Point

                                                                     Figure 3. Section View of Test Grid in the BTLab

Extended Abstract                                                                                     Page 3 of 10                                                                    8 August 2008
For use by NEMI
      In accordance with ASHRAE Standard 113-2005 (ASHRAE 2005d), the ADPI values
were calculated for the “occupied zone” of the calibrated test facility that was configured to
simulate an open office area without partitions or furnishings. For the CAD configuration,
data from 1,008 locations within the occupied zone were analyzed. For the UFAD
configuration, two sets of data were analyzed: one that included the “clear zones” as defined
in ASHRAE 2005d (i.e., 1,008 locations within the occupied zone); the other that excluded
the clear zones by deleting the data from within two feet of the floor diffusers (i.e., 168 data
points were deleted). These clear zones are typically 1.5 – 3 feet in radius, which can
represent 15 to 30 percent of the interior occupied floor space.

Results and Discussion

      Results indicated that the ADPI values for both CAD and UFAD exceeded the comfort
threshold of 80% at full load conditions (86% at 30 Btu/h ft2), but that an inverted pattern of
ADPI values occurred at partial loads as shown in Fig. 5.

      Figure 5. ADPI Values and Local Air Velocity Ranges for CAD and UFAD Configurations.

Extended Abstract                       Page 4 of 10                             8 August 2008
For use by NEMI
      These results provide quantitative evidence that the ADPI procedures are valid for
selecting and placing the supply air diffusers during design (ASHRAE 2005c) and for
evaluating air distribution performance under carefully controlled conditions (ASHRAE
2005d). While a recommended procedure for selecting and placing supply air diffusers for
UFAD systems does not yet exist, these results also revealed that the ASHRAE 113-2005
(ASHRAE 2005d) procedure yields reliable data with which to calculate the ADPI and
evaluate air distribution performance for this configuration.

      As shown in Figure 5, the ADPI results revealed two distinct patterns of performance
that deserve further investigation.

     •   At full loads, the ADPI values for the CAD and UFAD configurations both
         exceeded the design threshold of 80% (i.e., 86%) for occupant acceptability.
     •   For partial loads in the CAD configuration, the ADPI value maximized at the
         intermediate load (i.e., 94%) and decreased at the minimum load but maintained a
         value exceeding 80% (i.e., 84%). This well-known pattern (Nevins 1973), which
         is also shown in Figure 5, has been utilized in the design of variable air volume
         (VAV) systems for several years (Straub 1986).
     •   For partial loads in the UFAD configuration, the standard and modified ADPI
         values minimized below the design threshold at the intermediate load (i.e., 56%)
         and increased but did not achieve the design threshold at minimum load (i.e.,
         63%). In this case, reducing the airflow rate to control for the reduced loads
         resulted in lower ADPIs at partial load conditions.

      To evaluate the effect of the clear zone, the ADPI results for the UFAD configuration
were recalculated, omitting the data within two feet of the diffuser (i.e., “modified” data set).
These results revealed that the modified data set had incrementally higher ADPI values with
increasing loads. The differences between the total and the modified data sets resulted in a
1.5 point increase in ADPI at minimum load, a 5.5 point increase at intermediate load, and
an 11.0 point increase at full load.

      The ADPI pattern for the UFAD configuration indicates that occupant discomfort is
more likely during partial load conditions, which occur most of the time. The cause or
causes for the “inverted pattern” and low ADPI values for the UFAD configuration are
likely to be related to the function of the induction zones as the air jets dissipate to a
“terminal velocity” within the occupied zone. As shown in Figure 5, these terminal
velocities did not dissipate to less than the threshold value of 70 fpm within the occupied
zone at intermediate and full loads, whereas the terminal velocities for the CAD
configuration dissipated to less than 70 fpm for all loads.

      Analysis of the temperature and air velocity profiles in Appendix G of the Final Report
(Tan et al 2008) revealed that a measureable stratification of approximately 2 °F occurred in
the UFAD configuration at planes 48 – 60 inches above the floor for minimal and
intermediate loads in the UFAD, but this stratification was not apparent at full load. Rather,
columnar differences occurred with approximately 2 °F lower air temperatures in the
columns, which rose to 84 inches above the floor. Comparable profiles for the CAD

Extended Abstract                       Page 5 of 10                              8 August 2008
For use by NEMI
configuration revealed no measureable vertical stratification or columnar effects for the
three loads.

      These data also revealed that in the UFAD configuration the lowest air temperatures
within the occupied zones occurred at locations of highest air velocities, and highest air
temperatures occurred at locations of lowest air velocities. This combination of factors
resulted in minimizing the resultant ADPI at partial loads. Conversely, in the CAD
configuration, more thermal mixing and smaller velocity ranges in the occupied zones were
apparent at partial loads than at full load, which maximized the resultant ADPI at partial
loads. That the differences between standard and modified ADPI values increased with
increasing loads also indicates that the “clear zone” effect was detected in the ADPI
procedure. This finding supports the use of ASHRAE 113-2005 for evaluation of UFAD air
diffusion performance.

      The data from the ADPI tests were also analyzed to compare the calculated heat
extraction rates with the measured heat gains for both configurations and for all three loads.
As shown in Figure 6, the heat extraction rates in the CAD configuration matched the heat
gains to within 3 Btu/h ft2, indicating recovery rates of 66% at minimum load to 93% at full
load. As the minimum heat gain was from the lights located in the ceiling and the
intermediate and full loads also included resistance heaters located on the floor, see Figure
1, these results indicate that convection was the primary mechanism in the CAD
configuration for heat dissipation from the radiant and convective heat sources (i.e., ceiling
lighting and floor resistance heaters).

       The heat extraction rates in the UFAD configuration were not as closely match to the
heat gains. At minimum and intermediate loads, the differences ranged from 4.6 – 7.2 Btu/h
ft2, and at full load the difference was as much as 11 Btu/h ft2. These differences indicate
recovery rates of only 35 – 64% and that other mechanisms of heat dissipation from the
radiant and convective heat sources are significant. These findings are consistent with
literature that reports up to 40% of the space cooling load in UFAD systems is transferred to
the supply air plenum (Bauman, 2007).

Extended Abstract                      Page 6 of 10                             8 August 2008
For use by NEMI
                     Figure 6 Measured Heat Gain and Heat Extraction Rates

Conclusions and Recommendations

      This comparative analysis has achieved the primary objective of demonstrating that the
ADPI evaluation method, as specified in ASHRAE 113-2005, provides valid and reliable
results for UFAD systems in an open room, under unoccupied conditions (see Figure 1).
This conclusion is based on analysis of the data for both CAD and UFAD configurations
obtained in a calibrated facility, and upon the close agreement between the literature and the
design and evaluation data for three loads in the CAD configuration.

      Of particular importance is that the ADPI procedure is applicable to evaluate the
performance of the total UFAD area, including the “clear zones,” in terms of the combined
effects of low air temperatures and high air velocities (e.g., draft discomfort) and high air
temperatures and low air velocities (e.g., staleness and odors) on occupants, especially
during partial load conditions. Because clear zones can account for 10 – 30% of the
occupied space, it is recommended that the ADPI method be used to evaluate the total area

Extended Abstract                      Page 7 of 10                             8 August 2008
For use by NEMI
and to identify means to improve UFAD performance by reducing draft and staleness

      With regard to the second objective, these results indicate that the ADPI method in
ASHRAE 113-2005 is useful under closely controlled, laboratory-type conditions.
However, it requires sophisticated instrumentation and is limited to open areas without
furniture or occupants. This method is adaptable to job-site field evaluations, but
modifications are required including: 1) methods to account for physical obstacles such as
enclosed offices, moveable partitions, and furniture; 2) the presence of occupants; 3) the
characterization and quantification of heat gains from actual sources in the spaces; 4) the
application of field-quality instrumentation; and 5) simplification of the rigorous
calculations that are now required in the standard. It is recommended that a cost-effective
field standard be developed and that it be closely linked to the development of a design
standard that accounts for the disturbances of occupied spaces when selecting and placing
supply and return air devices for both CAD and UFAD configurations.

      Although the throw and spread characteristics of swirl-type diffusers are available in
manufacturers’ literature, an equivalent method to the T50/L approach to selecting and
placing these diffusers to achieve a threshold of 80% or greater in a UFAD system has not
yet been developed. Rather, current practice is to select the number of diffusers based on
the calculated space cooling load and the associated supply airflow rate for an assumed
difference in temperature between the supply air at the diffuser and the return air at the grille
or plenum. As this analysis demonstrated, current methods to estimate the relationship
between the heat gains and heat extraction rates or the temperature differences between the
supply and return devices are not reliable. For the third objective, it is recommended that a
design method for selection and placement of swirl-type and other floor diffusers for UFAD
systems be developed that is based on an equivalent T50/L parameter, predictable heat
extraction rates, and validation with performance data from actual installations in the field.

    Overall, this comparative analysis demonstrated that the air diffusion performance of
UFAD systems is dependent on several additional factors than those traditionally considered
when selecting and placing the diffusers and grilles for CAD systems, including:
    1) The uniformity of the temperature and static pressure in the plenum that supplies
          the diffusers;
    2) The significant percentage of heat gain that is not dissipated by diffusion and
          convection through the supply air diffusers and return air grilles;
    3) The percentage of floor area that is defined by the “clear zones,” which have
          deviations of temperature and air speed that are known to cause thermal

      Based on the data analyzed in the BTLab, the ADPI for the UFAD configuration was
similar to that for CAD configuration at the full load condition, as significant mixing
occurred in both configurations. However, at partial loads, the temperature and air velocity
factors caused a profile inversion between the CAD and UFAD configurations and resulted
in ADPI values for the UFAD configuration below those for occupant satisfaction at partial
loads. As health and comfort are of primary concern in the design and operation of air

Extended Abstract                       Page 8 of 10                              8 August 2008
For use by NEMI
diffusion and convection systems, it is recommended that further research be focused on
achieving a minimum of comfort threshold of 80% with both CAD and UFAD
configurations, especially at part-load conditions.

The authors are thankful to the National Energy Management Institute who provided
funding for this comparative analysis. We thank Liangcai (Tom) Tan, Ph.D., Research
Associate at the University of Nevada, Las Vegas, for his dedication to the design of the
BTLab and to the quality of his UFAD research in that facility. We also thank Richard
Sweetser for his assistance in this analysis. Finally, we express our gratitude to the National
Center for Energy Management and Building Technologies for its continued support of this

ASHRAE Standard 55-2004. 2004. Thermal Environmental Conditions for Human
Occupancy, ASHRAE, Atlanta GA.

ASHRAE Handbook of Fundamentals. 2005a. Chapter 8: Thermal Comfort. ASHRAE,
Atlanta GA.

ASHRAE Handbook of Fundamentals. 2005b. Chapter 33: Space Air Diffusion.
ASHRAE, Atlanta GA.

ASHRAE Handbook of Fundamentals. 2005c. Chapter 13: Odors. ASHRAE, Atlanta GA.

ASHRAE Standard 113-2005. 2005d. Method for Testing Room Air Diffusion, ASHRAE,
Atlanta GA.

ASHRAE Standard 62.1-2007. 2007. Ventilation for Acceptable Indoor Air Quality,
ASHRAE, Atlanta GA.

Bauman FS. 2003. Underfloor Air Distribution (UFAD) Design Guide. ASHRAE , Atlanta GA.

Bauman F, Webster T, and Benedek C. 2007. Cooling airflow design calculations for
UFAD. ASHRAE Journal, 49 (10): 36-44.

Fanger PO. 1988. Introduction of the Olf and Decipol units to quantify air pollution perceived by
humans indoors and outdoors. Energy and Buildings, 12: 1-6.

Int-Hout D. 2004. Best Practices for Selecting Diffusers. ASHRAE Journal Supplement, 46(6):

Nevins RG. 1973. Air Diffusion Dynamics: Theory, Design and Applications. Business News
Publishing Company, Birmingham, Michigan.

Extended Abstract                         Page 9 of 10                               8 August 2008
For use by NEMI
Rohles FH, Woods JE, Morey PR. 1989. Indoor environmental acceptability: Development of a
rating scale. ASHRAE Transactions, 95(1): 23-27.

Straub HE. 1986. Terminal Requirements based on Designers’ Terminal Choices. ASHRAE
Transactions, 92 (Part 1B): 519-527.

Tan L, Landsberger BJ, and Novosel D. 2008. Underfloor Air Distribution - An Experimental
Comparison Of Air Diffusion Performance Between UFAD and Conventional Air Distribution
(CAD) Systems. Final Report NCEMBT-080801, National Center for Energy Management and
Building Technologies, Alexandria, Virginia.

 Woods, J.E., D.T. Braymen, R.W. Rasmussen, G.L. Reynolds, and G.M. Montag. 1986..
"Ventilation Requirements in Hospital Operating Rooms - Part 1: Control of Airborne Particles".
ASHRAE Transactions, 92 (Part 2A): 396-426.

Extended Abstract                        Page 10 of 10                              8 August 2008
For use by NEMI

Shared By: