1、2008 ASHRAE 91ABSTRACTEmpirical models of airflow output and power consump-tion were developed for series fan powered variable air volumeterminal units at typical operating pressures. Terminal unitswith 8 in. (203 mm) and 12 in. (304 mm) primary air inlets fromthree different manufacturers were eval
2、uated. Generalizedmodels were developed from the experimental data with coef-ficients varying by size and manufacturer.Fan power and airflow data were collected at downstreamstatic pressures of 0.25 w.g. (63 Pa). Upstream static pressuresranged from 0.1 to 2.0 in w.g. (25 to 498 Pa). Data werecollec
3、ted at four different primary air damper positions and atfour terminal unit fan speeds. Model variables included theRMS voltage entering the terminal unit fan, the inlet air differ-ential sensor pressure, and the upstream static pressure.In all but one of the VAV terminal units, the resultingmodels
4、of airflow and power had R2values greater than 0.98.For the remaining unit, a faulty motor had been installed andshipped in the unit which prevented proper operation of theSCR. These models can be applied to HVAC simulationprograms to model series fan powered VAV systems.INTRODUCTIONVariable Air Vol
5、ume (VAV) systems maintain comfortconditions by varying the volume of primary air that is deliv-ered to a space. VAV terminal units that include a fan toimprove circulation within a zone are called fan poweredterminal units. These terminal units can draw in return air fromthe plenum space and mix it
6、 with primary air from the centralAir Handling Unit (AHU).When the fan is in the path of the primary airflow, theconfiguration is called a series terminal unit (Figure 1). Duringnormal operations, the terminal unit fan usually remains onexcept during un-occupied times in the zone. The controllerwill
7、 modulate the terminal unit damper in response to thecontrol signals from the thermostat and the inlet air differentialsensor. The inlet air differential sensor within the primaryairstream allows the controller to maintain a consistent volumeof airflow to the zone depending on the temperature setpoi
8、nt.The fans on these terminal units output a constant amountof air that does not vary with load because the downstreampressure is constant (Alexander and Int-Hout 1998). As aresult, when the primary air damper closes, more plenum airis induced and recirculated into the space. When the signalfrom the
9、 air velocity sensor indicates that the primary airflowhas reached a predetermined minimum (because of ventilationrequirements), the damper will not close any more. If the spaceis still too cold, electric or hot water supplemental heat can beused to meet the thermostat setpoint. To allow for various
10、 fanairflows, the units are typically equipped with a siliconcontrolled rectifier (SCR) fan speed controller.There is a need to develop a better understanding ofsystems using parallel and series fan powered VAV terminalunits. To model a system properly, it is important to be able tocharacterize the
11、individual terminal units. To date, there hasbeen little work in this area.The primary goal for this research was the development ofempirical models of power and airflow output for series fanpowered terminal units at typical operating pressures. Threemanufacturers (labeled A, B, and C) provided seri
12、es terminalunits for this work. An experimental setup was developed andused to test the fan powered terminal units. An experimentalprotocol was developed and used for all tests. Statistical anal-yses of experimental data were performed and used to developPerformance of VAV Series Fan-Powered Termina
13、l Units: Experimental Results and ModelsJames C. Furr Dennis L. ONeal, PhD, PE Michael A. DavisFellow ASHRAE John A. Bryant, PhD, PE Andrew CramletMember ASHRAE Student Member ASHRAE James C. Furr is a thermal management engineer with Lockheed Martin, Fort Worth, Texas. Dennis L. ONeal is Holdredge/
14、Paul Professorand Head and Andrew Cramlet is a research assistant, Department of Mechanical Engineering, Texas A&M University, College Station,Texas. Michael A. Davis is a research engineer with and John A. Bryant is a visiting associate professor in the Department of MechanicalEngineering, Texas A&
15、M University Qatar, Doha Qatar.NY-08-014 (RP-1292)2008, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions, Volume 114, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either prin
16、t or digital form is not permitted without ASHRAEs prior written permission.92 ASHRAE Transactionsgeneralized models that can be applied to the different manu-facturers terminal units. The units included three 8 in.(203 mm) and three 12 in. (304 mm) units. Manufacturers A8 in. unit had the designati
17、on S8A, manufacturers B 12 in. unitwas S12B, etc.This paper is the third of three papers that describe thedevelopment of experimentally based models of VAV fanpowered terminal units. The first paper (Furr et al. 2008a)described the experimental setup and methodology used tomeasure the performance of
18、 parallel and series fan poweredunits. That paper also described the small differences betweenthe terminal units that included the rated power of the terminalunit fan, the style of the primary airflow damper, and the styleof the backdraft damper. In the second paper (Furr et al. 2008b),the performan
19、ce of six parallel fan powered terminal units fromthree manufacturers was measured and characterized.DATA ANALYSIS METHODOLOGYOne goal of this research was to determine if a singlegeneralized model could be used for all the series terminalunits tested for a given size. Because of design differences
20、inthe units, the performances of the same sized units varieddramatically. Thus, no single model could be used to describea given size unit. The models had the same form, but useddifferent coefficients for the different sizes and manufacturers. Variables were first identified that were expected to be
21、significant in explaining fan airflow and power. Models werethen developed by determining the most statistically influen-tial independent variables using the F statistic. The variablewith the largest F statistic was added first. This method ofadding terms to the model was continued until no other va
22、ri-ables added were significant, defined as when the variables Fstatistic was below 4.0. Between each step, models werecompared against each other according to their adjusted coef-ficient of determination, R2adj(Neter et al. 1996). In developing the models for the series units, several vari-ables we
23、re considered: the SCR voltage, inlet air differentialpressure (Piad),upstream pressure (Pup), and primary airflow(Qprimary). The models for all of the series terminal units werecompared against each other. Any differences in termsincluded in the airflow or power models were investigated inan effort
24、 to create a single form model that would be applicableto all of the terminal units.RESULTS AND MODELSFan Terminal Unit AirflowThe fans on the series units used centrifugal, forward-curved style fans. These fans were expected to follow typicalfan curves and fan laws (ASHRAE 2001). The SCR voltage,up
25、stream static pressure, and inlet air differential pressurewere expected to be variables that could influence the capacityof the terminal unit fan.SCR setting was a variable in the model that had to bequantified. Each SCR setting corresponded to a different fanspeed. A simple experiment was conducte
26、d to determine therelationship between SCR setting and the speed of the fan. Atachometer was instrumented to an 8 in. (203 mm) parallel fanterminal unit from manufacturer A. Because the same motorsand SCRs were used with the parallel and series terminal units,it was assumed that the relationship bet
27、ween SCR setting andfan speed would be the same for fans in both the series andparallel units. It would have been preferred to test the seriesunits fans, to verify this assumption. However, given that thefan in a series unit is inside the terminal unit, it would havebeen difficult to ensure a consta
28、nt pressure difference acrossthe fan so that the relationship between fan speed and SCRsetting could have measured.At several different voltage settings, the RPM of the fanwas measured. During this testing, the upstream and down-stream static pressures were maintained constant to eliminatethe effect
29、s of pressure on the fan speed. A quadratic equationwas fit to the data (Figure 2) and had a R2value of 0.999.This test was conducted on two other terminal units,parallel terminal units P12B and P8C, which resulted in R2values of 0.994 and 0.997, respectively. Because of the high R2Figure 1 Series V
30、AV fan-powered terminal unit.ASHRAE Transactions 93values for the variety of groups and sizes, it was assumed thata general quadratic relationship would remain true for all ofthe terminal units.According to the fan laws, there should be a linear rela-tionship between airflow and fan speed (ASHRAE 20
31、01).Because a quadratic equation had been used to show the rela-tionship between SCR voltage and fan speed, it was assumedthat an equation of the same form could be used for the rela-tionship between SCR voltage and fan airflow.After this relationship was established, the other factorsthat were cons
32、idered in the modeling of the air output of thefan were the pressures immediately upstream and downstreamof the fan. Because the downstream static pressure was main-tained at the same value for all tests, it was not used explicitlyas an explanatory variable for the model. Another pressure thatcould
33、influence the airflow output of the unit fan would be thepressure inside the terminal unit, immediately upstream of thefan, Punit(Figure 3).During normal operation, some air was always inducedinto the terminal unit. Thus, the static pressure within theseries terminal unit was always sub-atmospheric
34、but the pres-sure was not measured. In planning for the experiments, theredid not appear to be a good way to instrument the terminal unitto measure this pressure accurately. After statistical analysis,it was determined that the pressure that the inlet air differentialpressure, Piad, was a suitable v
35、ariable to include in the modelto estimate the influence of the internal terminal unit staticpressure. For example, when an airflow model using V, V2,Piad, and Pupwas regressed for the series terminal unit S8C,the resulting F statistics for Piadand Pupwere 160 and 15,respectively. Because both F val
36、ues were greater than 4.0 bothvariables could have been used in the model. However, themodel using only V, V2, and Piadfor the S8C terminal unitobtained an R2adj value of 0.989. This model was deemedsufficient and in an attempt to maintain model simplicity, thevariable Pupwas not included in the air
37、flow models for theseries units. The resulting model for predicting the airflow inseries terminal units was a function of the SCR voltage and theinlet air differential pressure.Five of the six series terminal units had very similarresults for outlet airflow as a function of inlet air differentialpre
38、ssure. Two samples are shown in Figures 4 and 5 for termi-nal units S8A and S12C, respectively. The gentle slopes of thelines indicate that airflow was only slightly dependent on Piad.These results support the premise, found in literature (Alex-ander and Int-Hout 1998), that variations of upstream d
39、uctpressure, primary airflow, and damper position have littleFigure 2 Effect of SCR voltage on fan speed for parallelterminal unit P8A.Figure 3 Series VAV fan-powered terminal unit withpressure measurement locations.Figure 4 Fan airflow for series terminal unit S8A.Figure 5 Fan airflow for series te
40、rminal unit S12C.94 ASHRAE Transactionseffect on the pressure inside a series terminal unit, resulting infairly constant airflow. After a series terminal unit has beenbalanced for airflow, the air output of the series terminal unitshould be relatively constant despite changes in the upstreamconditio
41、ns.The airflow results from series terminal unit S12Bshowed much more scatter than the results from the otherterminal units (Figure 6). After trouble shooting the unit anddiscussions with the manufacturer of this unit, the disparitywas due to an incorrect fan motor installed and shipped inthe unit.
42、This motor prevented the SCR from workingcorrectly: the full range of SCR settings on this unit onlyresulted in a difference of 30 V as compared to differences ofover 100 V in the other units. The result was that there wasno discernable distinction in airflow output for differentSCR settings.Analysi
43、s of the data from unit S12B showed that thequadratic relationship between the SCR voltage and fan outputwas not evident. After initially developing models thatincluded V2and V, the F statistics were 0.04 and 0.22, respec-tively. A model developed using only V resulted in an F statis-tic of 34. Incl
44、usion of the squared term was never significant.This was probably due to the SCR/motor combination notbehaving as the ones in the other terminal units that weretested.The fan terminal unit output airflow model in series fanterminal units is shown in Equation 1. The coefficients foreach unit are pres
45、ented in Table 1.(1)Equation 2 is the model to characterize series terminalunit S12B, which was determined to have a faulty motor. Inthis model, V captures the small effect that SCR setting hason the airflow output. Pupand Piadwere both included inthis model, because their F values in the model were
46、 88 and83. Table 2 provides the coefficients for the model of thisterminal unit.(2)Power ModelData analysis of the power data for each of the terminalunit fans revealed a common characteristic. In each of theunits, there appears to be a linear response between power andfan airflow. Figures 7 and 8 s
47、how data for S8A and S12C,respectively. Because of this linearity, the resulting model topredict power maintained the same form as the model for fanairflow, and was a function of the SCR voltage and the inlet airdifferential pressure.As mentioned in the previous section, terminal unit S12Bproduced i
48、nconsistent data from the others, possibly becauseof a malfunctioning SCR. Those results provided little differ-ence in the airflow for the various SCR settings. In the analysisof the power data for this terminal unit (Figure 9), there wasalso little difference in power for the various SCR settings.
49、The fan power model for series fan terminal units, Equa-tion 3, was a function of the SCR voltage and the inlet airTable 1. Airflow Model Coefficients forSeries Terminal UnitsNameC1, cfmC2, cfm/V2C3, cfm/VC4, cfm/in. w.g.R2adjS8A 1776 0.0228 16.49 0.0036 0.989S8B 1705 0.0254 18.15 0.0448 0.994S8C 1310 0.0183 13.94 0.0677 0.997S12A 778.5 0.0091 6.918 0.0394 0.993S12C 1903 0.0105 16.78 0.0812 0.990Table 2. Airflow Model Coefficients forSeries Terminal Unit S12BNameC5, cfmC6, cfm/V2C7, cfm/VC8, cfm/in. w.g.R2adjS12B 925.7 2.68 55.8 293.2 0.688Figure 6 Fan airflow for series
