1、350 2009 ASHRAEABSTRACTDedicated outdoor air systems (DOAS) are often thought to be inferior to all air systems when contaminant flushing rates from occupied spaces are considered. This is a particu-larly prevalent attitude when the parallel terminal sensible cooling equipment provides no, or minima
2、l, filtration of termi-nal air, such as the case with chilled ceilings, beams, and even fan coil units. The central thrust of this paper is to determine if the attitude concerning the perceived inferiority of DOASs contaminant flushing rate is warranted. The investigation is carried out via an analy
3、tical case study involving a multi zone facility served by either a DOAS or a Variable Air Volume (VAV) system.DOAS BRIEFLY DEFINEDA DOAS delivers 100% outdoor air (OA) to each individ-ual space in the building via its own duct system, at flow rates generally as dictated by ASHRAE Std. 62.1 or highe
4、r. Elevated design ventilation flow rates may be necessary for latent load control, building pressurization, or to garner LEED green building points. Based upon the requirements of ASHRAE Std. 90.1, most DOAS applications require the use of total energy recovery. As a general rule a DOAS operates at
5、 constant volume during all occupied hours.Consequently, for most applications, the DOAS is not capable of meeting all of the thermal loads in the space by itself, and requires a parallel system to accommodate any sensible and latent loads the DOAS cant accommodate. The DOAS is not to be confused wi
6、th what is commonly called a “100% OA system”, whose flow rate is selected to meet the entire building sensible and latent loads. In other words, a DOAS generally delivers only about 20% as much air to a space as a “100% OA system”.The thermodynamic state of the delivered air varies1, but as a minim
7、um it should condition the air to the desired space dew point temperature (DPT), thus decoupling much of the latent load from the parallel system charged with the bulk of the space sensible load control.From a contaminant transport point of view, the constant volume DOAS leads to predictable pressur
8、e differentials (including neutral if desirable) between adjoining spaces or zones, thus minimizing the potential for interzonal transfer of airborne contaminants. Also since it does not use any recircu-lated air, airborne contaminants that may be present in one zone are not immediately distributed
9、throughout a facility by the mechanical system, as is common with mixing air systems ( i . e . VAV ) . The selection of the parallel system based upon contam-inant transport is important for two main reasons: first the parallel system may or may not recirculate and filter air locally, and second the
10、 parallel system may recirculate and filter air centrally, such as an all-air VAV system. THE FACILITY AND ASSUMPTIONS FOR THE CASE STUDYConsider a 20,000 ft2(1,858 m2) facility with 10 ft (3 m) high ceilings, Figure 1, consisting of a 2 zone perimeter region; 1,000 ft2(93 m2) zone 1 and 9,000 ft2(8
11、36 m2) zone 2 respectively. The facility also has a large interior 10,000 ft2(929 m2) zone 3. For the sake of the study, the following air flow rates will be used in the analysis:Contaminant Transport and Filtration Issues with DOASStanley A. Mumma, PhD, PEFellow/Life Member ASHRAES.A. Mumma is prof
12、essor emeritus in the Department of Architectural Engineering, Penn State University, University Park, PA. LO-09-032 2009, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2009, vol. 115, part 2. For personal use only.
13、 Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.ASHRAE Transactions 351VAV system. Supply air (SA) flow rate, 16,000 cfm (7,550 l/s), of which 4,000 cfm (1,888 l/s) is OA. Perim-eter zones 1 and 2, each
14、 receives 1 cfm/ft2(5 l/s-m2) of supply air via a shut off box VAV system, Figure 2 (from here on just referred to as a VAV system). Since the SA is 25% OA, then each perimeter zone is receiving 0.25 cfm/ft2(1.3 l/s-m2) OA. Interior zone 3 receives 0.6 cfm/ft2(3 l/s-m2) of supply air via a VAV syste
15、m. That translates to 0.15 cfm/ft2(0.76 l/s-m2) OADOAS. OA flow for the facility, 4,000 cfm (1,888 l/s) uniformly distributed in each zone, Figure 3, or 0.2 cfm/ft2(1 l/s-m2).The analysis is based upon the following additional assumptions:Well mixed zones, i.e. uniform concentrations.No interzonal t
16、ransfer, i.e. neglect influence of pres-sure differentials, human activity and or infiltration/exfiltration.Contaminants stay suspended, i.e. they do not settle or plate out in the zone.VAV system is analyzed while operating in the mini-mum OA mode (4,000 cfm (1,888 l/s) OA), and at the design suppl
17、y airflow rate, i.e. 16,000 cfm (7,550 l/s). Contaminate releases during full economizer mode (resulting in very high peak space concentrations when releases occur near the OA inlet) will not be presented.The capacitance of the duct system, and its associated influence on the transient response, is
18、neglected.The governing equations for the VAV follow: and DOAS follow: Figure 1 Floor plan, zone labels, and areas. V1dC1dt- Q1CmC1()=Figure 2 Schematic for the VAV system.Figure 3 Schematic for the DOAS.V2dC2dt- Q2CmC2()=V3dC3dt- Q3CmC3()=CexhQ1C1Q2C2Q3C3+Q1Q2Q3+-=Cm1 f()QOACOAQrecCexh+Q1Q2Q3+-=Qre
19、cQ1Q2Q3+()QOA=V1dC1dt- Q1CinC1()=352 ASHRAE TransactionsWhere: = concentration of contaminant = SA flow rates =space volume= filter efficiencyCONTAMINANT INTRODUCTION NEAR THE OA INTAKEOnly contaminant sources near the OA intake are presented. Contaminant transport from internal sources2will not be
20、presented since filtration has no bearing on concentra-tions for DOAS. The contaminant source is assumed to be near the OA stream and of 5 minute duration. Filter efficiency can significantly impact the peak concentrations and transient response of the zone concentra-tions when the contaminant sourc
21、e is near the OA intake. First consider the case where the filters have no impact on contaminant removal, i.e. zero percent filter efficiency. Figure 4 illustrates such a case with the release near the OA intake. Since the VAV supply air quantity to the exterior zones 1 and 2 is higher than that of
22、the interior zone 3, their peak concen-trations are different. For the DOAS system, since the supply airflow rate is a constant 0.2 cfm/ft2(1 l/s-m2) the peak concentration in all zones and the subsequent response is uniform throughout the facility. Note: the zone transient concentrations are presen
23、ted in dimensionless form, refer-enced in all cases to the peak concentration experienced by the DOAS system with zero percent filtration efficiency. Since the VAV supply airflow rates differ between the interior and perimeter zones, their peak concentrations differ (with the higher flow/unit floor
24、area perimeter zones OA/ unit floor area exceeds that of a DOAS 20% and the interior zone OA/ unit floor area is 20% lower than DOAS). The higher VAV supply airflow rate to the zones 1 and 2 not only causes their peak concentrations to be higher than that of the interior zone 3, but also causes them
25、 to clear faster than the interior zone 3.Typically it is assumed that the filtration efficiency used with a VAV system and a DOAS system are equal. That need not be true. And if the filtration efficiency for a particular contaminant is 80% for a VAV system, the zone concentrations for both the VAV
26、and DOAS systems are equal, exactly one hour following the release, if the DOAS filter efficiency is 98% as illustrated in Figure 5. The area under the curves repre-sents the exposure that the occupants of the zones would expe-rience. It is significant to note that with the improved efficiency DOAS
27、filtration, the integrated one hour exposure immediately following the release is about 425% more in the facility served by the VAV system than that the a DOAS system. This is true since the 98% filter in the DOAS OA path removes 98% of the material before delivering it to the space. V2dC2dt- Q2CinC
28、2()=V3dC3dt- Q3CinC3()=Cin1 f()COA=CQVfFigure 4 Transient dimensionless concentration from release near OA inlet with filter efficiencies of 0%.Figure 5 Transient response to a release nears the OA intake, VAV filters efficiency. 80%, DOAS filter efficiency 98% for same space concentration after 60
29、min.ASHRAE Transactions 353So even though the DOAS spaces clear slower than the VAV system, its peak concentration is only 2% what it experiences without a filter.For the facility used in this analysis, employing the asso-ciated assumptions and transient equations, it is easy to compute the DOAS fil
30、tration efficiencies necessary to match the lowest 60 minute VAV zone concentration and exposure. Those results are presented in Figure 6. Notice the VAV and DOAS filtration efficiencies match at zero and 100 percent. Otherwise the DOAS filter efficiency must be higher than the VAV for equivalence.
31、Finally, the DOAS filter efficiency necessary to match 60 minute exposure is less than that neces-sary to match the 60 minute concentration. OPTIMIZED FILTER SELECTION FOR A DOAS TO PROVIDE EQUIVALENT ONE HOUR EXPOSURE, OR SPACE CONCENTRATION, RELATIVE TO AN OPTIMIZED VAV FILTER EFFICIENCY SELECTION
32、Figures 5 and 6 illustrated that selecting a more efficient filter for a DOAS, when compared to a lower efficiency VAV, leads to similar exposures or interior concentrations when dealing with contaminant sources in the OA stream. That could lead one to theorize that the higher first cost per unit ar
33、ea and pressure drop of the DOAS filter would make such a selec-tion economically impractical. In an effort to test that hypoth-esis for the facility used in the case study, assuming that the average contaminant particle size is 2 micron, cost and oper-ating data for various efficiency filters was u
34、ndertaken and used in an optimization set of equations. The data presented in this paper is limited to the filter efficiency combinations presented in Table 1. Images of the filters used for the simu-lations are in Figure 7.Detailed performance data for commercially available 2 micron efficiency fil
35、ters were obtained from the manufactur-ers3, including clean and loaded pressure drop at a given face velocity, pressure drop as a function of loading, pressure drop as a function of face velocity, and recommended filter change frequency. These data were curve fit for use in the optimization solutio
36、ns.For the base case optimizations, the following assump-tions were made:1. VAV SA flow: 16,000 cfm (7,550 l/s); and DOAS SA flow: 4,000 cfm (1,888 l/s).2. Assume that the filter loading profile is exponential, causing the average pressure drop over the life of the filter to be: . Use 0.87 as the .3
37、. Reference replacement frequency is 3 months for VAV and 6 months for DOAS4. fan/motor combined efficiency: 60%.5. Electrical cost: $0.10/kWh.6. Term of analysis: 5 years.7. Fan operating hours per year: 4160 hours.8. Neglect the time value of money and inflation.Figure 6 DOAS filter efficiencies n
38、ecessary to match the space concentration or exposure after 60 min compared to the VAV filter efficiency, for occurrences near the OA intake. ploaded optpclean opt()2FillratioFillratioTable 1. VAV/DOAS Filter Efficiencies for Equivalent PerformanceNominal VAV Filter Efficiency, Merv RatingNominal DO
39、AS Filter Efficiency, Merv RatingFilter Efficiency for Equivalent 1 h Concentration/Exposure Performance (Figure 6)75, merv 11 94 97/8994, merv 13 98 99/9898, merv 14 99.7, merv 16 99.7/99.3354 ASHRAE Transactions9. Filter cost data was obtained from an online University Business Services purchasing
40、 contract4. Labor costs for filter replacement was not considered. 10. Op cost for parallel equipment (I.e. FCUs, Chilled Ceil-ings etc.) used with the DOAS have not been included in the analysis.The simple optimization equations solved5are:1.2.3.4.5.6.7.8.9.10.11.12.13.Where: Filter pressure drop w
41、hen clean at optimal face velocity, in-wg (Pa): Filter pressure drop when loaded at opti-mal face velocity, in-wg (Pa): Optimal face velocity, ft/min (m/s): optimized filter size, ft2(m2): accounts for the non-linear exponential filter loading characteristics: the average pressure drop over the life
42、 of the filter at the optimum conditions, in-wg (Pa): assumed combined fan/motor efficiency, fraction: Fan horsepower at the optimum conditions using the average filter pressure differential, HP: Fan energy demand, kW: ratio of filter pressure drop increase from clean to loaded at optimum FV vs. fil
43、ter pressure drop increase from clean to loaded at 500 fpm (2.5 m/s) ref FV. NOTE, this is used to alter the normal filter replace-ment period based on filter size (if (2.5 m/s) replacement is more frequent, and if (2.5 m/s), replacement is less frequent): months between filter changes with optimize
44、d filter selection.: first cost of the filter at optimized condi-tions.: fan operating cost as a result of the filter alone: operating hours per year: years in the economic analysis: utility cost, $/kWh: value of the objective equation, $The results of the base case optimizations are presented in Ta
45、ble 2.Perturbations to the base case were made including: cutting the electric rates or hours of operation by 50%, chang-ing the DOAS filters every month, reducing the fan motor effi-ciency, and filter average pressure drop with a fill factor of 1.0. These perturbations did not change the trends pre
46、sented in Table 2. One sizable perturbation was to force a DOAS filter change each month, greatly increasing the first cost of the filter driving its size down at the expense of operating cost. Those optimization results are presented in Table 3.DISCUSSION OF RESULTSFor the 20,000 ft2(1,858 m2) faci
47、lity with filter efficiency equal to zero for a particular contaminant, the transient Figure 7 Images of the filter types used.pclean opt c1FVoptc2FVopt2+=ploaded opt c3FVoptc4FVopt2+=FVoptSA FAopt=Fillratio0.87=pavg opt ploaded opt pclean opt()2Fillratio=Faneff0.6=FanHP optSA pavg opt()Faneff()=Fan
48、kW optFanHP opt0.75=pratioploaded opt pclean opt()( ploaded Manf ref FV-pclean Manf ref FV=Monthsbetween filter changes opt=Month sbetween filter changes opt ref FVpratio()Filter1st t optcosFAoptFilter $per unit area 12 Monthsbetween filter changes opt()term of analysis=Opt optcosFankW optOphours/yrPeriodyearselec $per kWh=Minimize Filter1st cost optOpcost opt+=pclean optploaded optFVoptFAoptFillratiopavg optFaneffFanHP optFankW optpratioFV 500 fpmFV 500 fpmMonthsbetween filter changes optFilter1st cost optOpcost optOphours/yrPeriodyearselec$per kWhMinimizeASHRAE Transactions
