Characterizing airflow and power of VAV series fan-powered terminal units from component data--part I.
A key component in variable-air-volume (VAV) systems is the air terminal unit, which is used to control the amount of conditioned air introduced into a space. Figure 1 shows the airflow in a VAV system. Primary air is conditioned by a central air-handler unit (AHU) and delivered by a central supply fan through a single supply air duct system to terminal units. By mixing primary and plenum air within terminal units, supply air is produced and provided to different zones. Terminal units using fans to regulate airflow are called fan-powered terminal units (FPTUs). They are typically installed in return plenums and ducted to air outlets. Compared with terminal units without fans, FPTUs offer several advantages, including reducing or eliminating reheat by mixing primary air with warmer plenum air, reducing central supply fan operating pressure and energy consumption in an air distribution system, and boosting downstream air pressure that allows the air to be delivered to the zones with insufficient airflow (ASHRAE 2012).
When a terminal unit fan is in series with primary air, this unit is referred to as a series FPTU. In a series FPTU, a terminal unit fan must be operating continuously whenever a central supply fan is on and all the air delivered to a conditioned zone passes through the terminal unit fan, as shown in Figure 2. When cooling load decreases, the inlet damper gradually closes to regulate primary air until it reaches its predetermined minimum position. At the same time, the fan induces more return air from the plenum to compensate for reduced primary air. By adjusting the amount of primary and plenum air, the temperature setpoint can be maintained without reducing the amount of supply air. An optional electric strip heater or hot-water coil can be used to provide supplemental heat.
Currently, the two types of motors used in series FPTUs are permanent-split capacitor (PSC) motors and electronically commutated motors (ECM). PSC motors are driven by alternating current (AC) and equipped with silicon-controlled rectifiers for motor speed control. ECM motors are brushless direct current (DC) motors with permanent magnet rotors and ball bearings (Roth et al. 2004). The torque-driven design of ECM motors allows them to maintain a relatively constant flow rate in an air terminal unit, regardless of downstream static pressures (Kenty 2007). The Nonresidential Compliance Manual for California's 2005 Energy Efficiency Standards has mandated the use of ECM motors in series FPTUs unless standard motors that can be shown to be at least 70% efficient are used (CEC 2005). To model a VAV system properly, it is important to characterize the performance of terminal units.
Recently, Furr et al. (2008a, 2008b, and 2008c) and Edmondson et al. (2011a and 2011b) have experimentally studied FPTUs. They utilized a system-oriented model to estimate airflow, pressure, and power of FPTUs with PSC and ECM motors. Because of the complexity in aerodynamic conditions and airflow patterns within FPTUs, they treated FPTUs as "black boxes" and mainly focused on correlating the airflow and power measurement with peripheral pressure settings. In their experimental setup, FPTUs were connected to two airflow chambers for the measurement of primary and fan airflow under varying upstream static pressures from 0.1 to 2.0 in. wg (24.9 to 498.3 Pa) and fixed downstream static pressures at 0.25 in. wg (62.3 Pa). Their models were limited by the black box approach and required knowledge of pressure conditions upstream and downstream of an FPTU.
The public domain building energy simulation program, EnergyPlus, uses a simplified component approach to simulate the performance of series and parallel FPTUs. FPTUs are treated as compound components, including a zone mixer, a constant volume fan, and a heating coil, which can use hot water, electric, or gas as input (EnergyPlus 2012a). Figure 3 shows the schematic of series and parallel FPTUs simulated in EnergyPlus (EnergyPlus 2012b). The zone mixer receives airstreams from multiple inlets and mixes them together. The simulation of the mixer involves simple mass and energy balances of airstreams. The model of constant-volume fan assumes the fan operates continuously based on a time schedule. To appropriately characterize the fan performance, the knowledge of overall fan efficiency, pressure rise, maximum flow rate, and motor position are required. The heating coil is also modeled with a simple mass and energy balances to estimate the amount of sensible heating required. The overall performance of FPTUs is estimated by simulating the three subcomponents in sequence. From the zone simulation results, EnergyPlus calculates the heating/cooling demand on FPTUs. If a zone needs cooling and the terminal unit is not scheduled off, then it is available to serve the zone. The outlet air temperature required to meet the zone load is calculated, and then the primary and plenum airflow rates can be determined. The models of fan and heating coil are used to calculate the energy use for each component.
Compared with the experimentally determined empirical models developed by Furr et al. (2008b and 2008c) and Edmondson et al. (2011a and 2011b), the component-based approach can be generally applied on any FPTU. However, it fails to capture the effects caused by air pressures either on the primary or supply side of the system. For example, both Furr et al. (2008b) and Edmondson et al. (2011a) found significant leakage caused by pressure differentials across parallel FPTUs. This leakage could dramatically increase the energy use of parallel FPTUs. Because pressure input is not included in the EnergyPlus model, leakage is not explicitly modeled. Thus, EnergyPlus can be expected to underpredict the energy consumption of a VAV system using parallel FPTUs. In addition, fan efficiency and power consumption also vary with inlet and outlet pressures, but these inputs are currently constant in EnergyPlus.
An important question is whether piecing together the individual components of FPTUs will provide adequate representation of overall system performance. If the individual components, when combined, do not provide a good performance representation of FPTUs, then the energy use estimated by building energy simulation programs may not be accurate.
The objective of this study was to investigate how well the performance of a series FPTU could be modeled by combining the models of its individual components, namely fan/motor/ controller, damper, and housing. The test procedure was developed and the experimental data for the individual components from eight series FPTUs were collected. Component models were established from the collected data, and then a system model was built by assembling the component models. In addition, the model outputs were compared with the experimental results from previous study (Edmondson et al. 2011b) for model verification. Since the developed models only focused on the air side and power performance, the evaluation of heating performance was not included in the scope of this study and was not conducted.
This is Part I of this study. In this paper, the experimental setup for fan testing was described and the models of fan airflow and fan/motor/controller power consumption were developed from the measured data. Additionally, overall fan efficiency and power factor were measured and reported. Part II summarized the experimental results and empirical models of primary and plenum airflow. It also included the development of a system model of series FPTU and model verification.
EXPERIMENTAL SETUP AND PROCEDURE
Eight series FPTUs from three manufacturers; labeled as A, B, and C; were experimentally studied. They were the same units used by Edmondson et al. (2011b). All units were equipped with ECM motors. They differed in the size of primary air inlet. Four of the units had 8 in. (203 mm) inlets and the other four had 12 in. (304 mm) inlets. The terminal unit with an 8 in. (203 mm) inlet from Manufacturer A was designated as S8A, and likewise Manufacturer B's 12 in. (304 mm) terminal unit was S12B, etc. Fans were named after the corresponding terminal units. For example, F_S8A was the fan from terminal unit S8A, while F_S12B was the name of the fan from terminal unit S12B. Table 1 summarizes the key characteristics of series FPTUs and fans tested in this study.
The experimental apparatus included a fan/motor/controller from a series FPTU, a supply duct, and an airflow chamber, as shown in Figure 4. Fan/motor/controllers were removed from each series FPTU and directly connected to the airflow chamber through a rectangle sheet metal duct with a length of 3.5 hydraulic diameters. This duct had the same cross-sectional dimensions as the fan outlet and was constructed in accordance with ASHRAE Standard 130 (1996).
Airflow was measured by using the airflow chamber. This chamber had a nozzle board consisting of one 1 in. (25 mm), one 3 in. (76 mm), and four 5 in. (127 mm) nozzles. It was built to ASHRAE Standard 51 (ASHRAE 2007) requirements for an outlet chamber setup. An assist blower controlled by a variable-frequency drive (VFD) was attached to the chamber and was used to adjust discharge static pressure. Airflow through the known open nozzle areas was calculated based on the measurement of discharge static pressure and nozzle differential pressure. The calculation procedure was directly adopted from ASHRAE Standard 51 (ASHRAE 2007). Raw airflow data were adjusted to standard conditions of temperature and barometric pressure to compensate for environmental changes in the period of data collection.
A stand-alone psychrometric station was used to monitor the temperature and humidity of ambient air. Air density was calculated based on this measurement in conjunction with barometric pressure. Air pressures were measured by pressure transducers with 4-20 mA output. Fan electrical performance was evaluated using a power quality analyzer. The simultaneously measured and recorded data included real and apparent power, root mean square (RMS) voltage and current, as well as power factor. The data files were cached in the power quality analyzer's internal memory and then downloaded to a PC. The instrument specification is listed in Table 2.
The outlet airflow chamber was verified over a wide range of flow rates from 150 to 2000 [ft.sup.3]/min (0.07 to 0.94 [m.sup.3]/s) by using another airflow chamber which was set up as an inlet chamber in accordance with ASHRAE Standard 51 (ASHRAE 2007). The two chambers were directly connected via a piece of duct. The volumetric flow rate obtained by the outlet chamber was compared with the value measured by the inlet chamber. Various nozzle combinations were used, depending on the flowrate investigated. In all cases, the differences between the two chambers did not exceed [+ or -] 3%.
Fan speed and discharge static pressure were key parameters influencing fan performance. Variable fan speed was realized by using the ECM controller that came with each fan/ motor. Two types of controllers were encountered. Manufacturer A provided a controller that consisted of a LCD display and a potentiometer. It could be set to a numerical value between 0 and 100 to represent fan speed. The other type of controller used DC voltage as input signal. Manufacturers B and C employed this design. Each manufacturer also had different design operating ranges for their series FPTUs. According to the catalog data provided by each manufacturer, a test matrix spanning these two independent variables was established to cover the expected operating ranges in the field. Table 3 shows the independent variables and their ranges for fan testing. At the specified ECM setting and discharge static pressure, the corresponding airflow and electrical performance were measured.
RESULTS AND MODELS
Experimental data were collected on airflow and power consumption over a range of fan speeds and discharge static pressures. The corresponding models were developed from the measurement.
Figure 5 shows the airflow plotted against ECM setting for F_S8A. At a given ECM setting, the airflow obtained at different discharge static pressures were almost the same. These data indicated that the fans with ECMs were able to provide relatively constant airflow regardless of changes in discharge static pressure. ECMs were designed to sense the motor's torque changes and automatically match the output torque to the torque required to maintain motor's setpoint. Thus, variations in discharge static pressures would not significantly impact the fan airflow. Moreover, Figure 5 also demonstrated that fan airflow was mainly a function of ECM setting and a linear relationship existed between these two variables. This observation confirmed the result reported by Edmondson et al. (2011b), who found that airflow delivered by ECM fans was primarily dependent on fan speed control signal and showed less dependence on inlet air velocity pressure.
ECM setting was chosen as the only input variable for the fan airflow model. Considering the variety in ECM controller input, a single variable was needed to describe the ECM setting for all the fans from different manufacturers. A dimensionless variable p was defined. It represented the percentage of input signal to the controller over the whole operating range.
A linear regression was conducted and Equation 1 was used to fit the experimental data.
[[??].sub.fan] = [C.sub.1] + [C.sub.2] x [eta] (1)
[C.sub.1] and [C.sub.2] are empirical parameters estimated from the experimental data. They varied by terminal units. Figure 6 compares the model prediction with experimental measurement for fan airflow. The predicted and measured values were in close agreement with each other. The error was bounded within [+ or -] 10% of measurement.
The empirical parameters and coefficient of determination ([R.sup.2]) for developed model were reported in Table 4. All [R.sup.2] values were above 0.99. The agreement shown in Figure 6 and high [R.sup.2] values listed in Table 4 provided evidence that the model was adequate to characterize the fan airflow performance.
Figure 7 shows the fan-power consumption plotted against discharge static pressure for F_S8A. At a given ECM setting, fan power increased linearly when discharge static pressure was increased. At a given discharge static pressure, an increase in ECM setting always led to an increase in fan-power draw. The similar trend was also observed for the other fans. The two most important factors affecting fan power were ECM setting and discharge static pressure. So these two variables were used as the input variables for the fan-power model. The format of this model is shown in Equation 2.
[Power.sub.fan] = [C.sub.1] x [eta] + [C.sub.2] x [[eta].sup.2] + [C.sub.3] X [P.sub.discharge] (2)
[C.sub.1], [C.sub.2], and [C.sub.3] are empirical parameters determined from the experimental data. A comparison between estimated and measured fan-power consumption for all the eight fans was shown in Figure 8. The prediction from the developed model closely matched the experimental data. The empirical parameters and [R.sup.2] values for developed fan-power model were listed in Table 5.
The total and static efficiencies of each fan/motor combination were calculated from the experimental data. The total efficiency is defined as the ratio of fan/motor output to the fan/ motor power input, as shown in Equation 3. The fan/motor static efficiency was determined from the total efficiency and the ratio of static pressure to total pressure using Equation 4. Both equations were adopted from ASHRAE Standard 51 (ASHRAE 2007). In these equations, [P.sub.t] and [P.sub.s] are the total and static pressure changes across a fan. Since fans were tested in the configuration of open inlet in this study, the total pressure at the fan inlet was considered equal to 0. The fan total pressure change was numerically equal to the total pressure at the fan outlet, which is calculated from Equation 5. [P.sub.v] is velocity pressure and calculated from Equation 6.
[[eta].sub.t] = [Q.sub.fan][P.sub.t]/[Power.sub.fan] (3)
[[eta].sub.s] = [[eta].sub.t]([P.sub.s]/[P.sub.t]) (4)
[P.sub.t] = [P.sub.v] + [P.sub.s] (5)
[P.sub.v] = [[rho]/2] x [([Q.sub.fan]/A).sup.2] (6)
Although fan/motor combinations from different series FPTUs did not perform the same, they did share some common characteristics. Figure 9 shows the fan/motor total efficiency for F_S8A. For a given discharge static pressure, the total efficiency increased following an increase in airflow. At 0.1 in. wg (24.9 Pa), total efficiency increased from 28% to 33% when the airflow increased from 390 to 1000 [ft.sup.3]/min (0.18 to 0.47 [m.sup.3]/s). The same trend was also observed in the other pressure settings. For a given airflow rate, the increase in discharge static pressure has a positive impact on total efficiency. For example, at the airflow of 1000 [ft.sup.3]/min (0.47 [m.sup.3]/ s), the total efficiency increased from 33% to 39% as the pressure was increased from 0.1 in. wg to 0.6 in. wg (24.9 to 149.5 Pa).
Figure 10 shows the fan/motor static efficiency for F_S8A. The static efficiency increased with increasing discharge static pressure and decreased with increasing airflow. For a given airflow rate, both fan/motor total and static efficiency increased with increasing discharge static pressure. However, for a given static pressure, the fan/motor static efficiency decreased dramatically when fan airflow rate increased. Overall, the eight fan/motor combinations had total efficiencies ranging from 11% to 50% and static efficiencies ranging from 1% to 22% at their maximum speed.
Both Figures 9 and 10 show that fan/motor efficiencies vary significantly with discharge static pressure and airflow rate. The data for all eight fan/motor combinations showed an even wider variation. These plots and data for the remaining fan/motor combinations show the difficulty of picking a single value to represent the fan efficiency, as required in EnergyPlus (2012a, 2012b).
Power factor is defined as the ratio of the real to apparent power. Figure 11 shows the power factor for F_S8A. The power factor increased responding to increasing discharge static pressure. Also, when the fan was operating at higher speeds, the power factor tended to increase. From the test results of the eight fan/motor combinations, the power factor varied between 0.27 and 0.56, which was consistent with the previous measurement (Edmondson et al. 2011b)
SUMMARY AND CONCLUSIONS
In the previous studies of Furr et al. (2008c) and Edmondson et al. (2011b), they developed detailed semi-empirical performance models for series FPTUs with PSC and ECM motors, respectively. However, due to the complexity in aero dynamic conditions and flow patterns within series FPTUs, they applied the black box approach and simply correlated the airflow and power performance with the measurement of peripheral pressures. This approach limited their models that were only applicable to certain units at specific working conditions. In this study, the component approach was taken. The ultimate goal of this research is to characterize the performance of series FPTUs by combining the models of individual components, namely fan/motor/controller, damper, and housing.
Eight fan/motor/controllers were removed from the series FPTUs used by Edmondson et al. (2011b) and tested as stand-alone units. The models of airflow and power consumption were generated from the measured data. In the airflow model, the nondimensional ECM setting was taken as the single input since the ECM fan was able to provide constant airflow regardless of pressure changes. The [R.sup.2] values of this model were all above 0.99. The fan/motor power consumption model also correlated well with measured data and had [R.sup.2] values ranging from 0.976 to 0.995. Comparisons were made between predicted and measured data in airflow and power consumption. The results indicated that the prediction using developed models can be as accurate as [+ or -] 10% of measurement.
Estimations of fan airflow and power consumption are required to characterize the performance of series FPTUs in the component-based modeling approach. Models produced in this paper can be used to predict the airflow and power performance of fans with ECM motors in series FPTUs, and thus can be used as the submodels to evaluate the whole terminal unit performance. This is the first part of this study. In Part II, the experimental results and empirical models of primary and plenum airflow will be introduced. In addition, the development of a system model of series FPTU using component-based modeling approach will be discussed. Model validation against experimental data will also be presented.
NOMENCLATURE A = cross-sectional area of duct at fan outlet, [ft.sup.2] ([m.sup.2]) [P.sub.discharge] = discharge static pressure, in. wg (Pa) [Power.sub.fan] = power consumption of terminal unit fan/motor, W [P.sub.s] = fan static pressure, in. wg (Pa) [P.sub.t] = fan total pressure, in. wg (Pa) [P.sub.v] = fan velocity pressure, in. wg (Pa) [Q.sub.fan] = amount of airflow through terminal unit fan, [ft.sup.3]/min ([m.sup.3]/s) [eta] = percentage of the controller voltage over the whole operating range [[eta].sub.t] = fan/motor total efficiency, % [[eta].sub.s] = fan/motor static efficiency, % [rho] = air density, [lb.sub.m]/[ft.sup.3] (kg/[m.sup.3])
ASHRAE. 1996. ANSI/ASHRAE Standard 130-1996, Methods of Testing Air Terminal Units. Atlanta: ASHRAE.
ASHRAE. 2007. ANSI/ASHRAE Standard 51-2007, Laboratory Methods of Testing Fans for Certified Aerodynamic Performance Rating. Atlanta: ASHRAE.
ASHRAE. 2012. ASHRAE Handbook--HVAC Systems and Equipment. Atlanta: ASHRAE.
CEC. 2005. Nonresidential Compliance Manual for California's 2005 Energy Efficiency Standards. Sacramento: California Energy Commission.
Edmondson, J., D.L. O'Neal, J.A. Bryant, and M.A. Davis. 2011a. Performance of parallel fan-powered terminal units with electronically commutated motors. ASHRAE Transactions 117(2): 885-93.
Edmondson, J., D.L. O'Neal, J.A. Bryant, and M.A. Davis. 2011b. Performance of series fan-powered terminal units with electronically commutated motors. ASHRAE Transactions 117(2): 876-84.
EnergyPlus. 2012a. EnergyPlus Engineering Reference: The Reference to EnergyPlus Calculations. Berkeley: Lawrence Berkeley National Laboratory.
EnergyPlus. 2012b. Input Output Reference: The Encyclopedic Reference to EnergyPlus Input and Output. Berkeley: Lawrence Berkeley National Laboratory.
Furr, J., D.L. O'Neal, M. Davis, J.A. Bryant, and A. Cramlet. 2008a. Performance of VAV fan-powered terminal units: Experimental setup and methodology. ASHRAE Transactions 114(1): 75-82.
Furr, J., D.L. O'Neal, M. Davis, J.A. Bryant, and A. Cramlet. 2008b. Performance of VAV parallel fan-powered terminal units: Experimental results and models. ASHRAE Transactions 114(1): 83-90.
Furr, J., D.L. O'Neal, M. Davis, J.A. Bryant, and A. Cramlet. 2008c. Performance of VAV series fan-powered terminal units: Experimental results and models. ASHRAE Transactions 114(1): 91-7.
Kenty, K.L. 2007. Use of electronically commutated motors (ECMs) in air terminal unit. ASHRAE Transactions 113(1): 334-41.
Roth, K.W., A. Chertok, J.Dieckmann, and J.Brodrick. 2004. Electronically commutated permanent magnet motors. ASHRAE Journal 46(3): 75-76.
Student Member ASHRAE
Dennis L. O'Neal, PhD, PE
Peng Yin is a PhD student in Mechanical Engineering at Texas A&M University, College Station, TX. Dennis L. O'Neal is a professor and dean of Engineering and Computer Science at Baylor University, Waco, TX.
Table 1. Series FPTU and Fan Characteristics Fan Fan Motor Fan Airflow Range, Horsepower, [ft.sup.3]/min hp (W) ([m.sup.3]/s) F_S8A 1/2 (373) 300-1050 (0.14-0.50) F_S12A 1 (746) 700-2500 (0.33-1.18) F_S8B 1/2 (373) 200-900 (0.09-0.42) F_S12B 1/2 (373) 400-1600 (0.19-0.76) F_S8C_M1 1/2 (373) 200-1100 (0.09-0.52) F_S8C_M2 1/2 (373) 200-1100 (0.09-0.52) F_S12C_M1 3/4 (559) 500-2000 (0.24-0.94) F_S12C_M2 3/4 (559) 500-2000 (0.24-0.94) Table 2. Instrument Specification Measurement Point Sensor Specification Ambient 0[degrees]F-100[degrees]F temperature (-17.8[degrees]C-37.8[degrees]C), [+ or -] 0.7[degrees]F ([+ or -] 0.39[degrees]C) Ambient relative 0-100%, [+ or -] 2% humidity Chamber static 0-10 in. wg (0-2492 Pa), pressure [+ or -] 0.25% full scale Chamber 0-6 in. wg (0-1495 Pa), differential [+ or -] 0.25% full scale pressure Voltage [+ or -] 0.1% of reading Current [+ or -] 1% of reading Power [+ or -] 1% of reading Table 3. Fan Test Matrix FPTU ECM Discharge Manufacturer Settings Static Pressure A 20%, 40%, 60%, 80%, 0.1-0.6 in. wg 100% full scale (24.9-149.5 Pa) B 4V, 6V, 8V, 10V 0.1-0.5 in. wg (24.9-124.6 Pa) C 2V, 4,V, 6V, 8V, 10V 0.0-0.5 in. wg (0-124.6 Pa) Table 4. Empirical Parameters and [R.sup.2] Value for Fan Airflow Model Fan [C.sub.1] [C.sub.2] [R.sup.2] F_S8A -19.258 10.110 0.998 F_S12A 73.093 22.607 0.996 F_S8B 60.233 15.119 0.994 F_S12B 420.260 12.639 0.994 F_S8C_M1 70.961 14.810 0.998 F_S8C_M2 24.375 12.974 0.990 F_S12C_M1 441.522 17.467 0.994 F_S12C_M2 32.695 21.458 0.995 Table 5. Empirical Parameter and [R.sup.2] Value for the Fan-Power Model Fan [C.sub.1] [C.sub.2] [C.sub.3] [R.sup.2] F_S8A -1.631 0.0341 199.283 0.985 F_S12A -5.760 0.151 316.009 0.995 F_S8B -3.041 0.0814 275.299 0.990 F_S12B -1.181 0.0533 272.298 0.993 F_S8C_M1 -1.673 0.104 160.150 0.995 F_S8C_M2 -1.035 0.0459 152.260 0.976 F_S12C_M1 2.102 0.0289 229.583 0.977 F_S12C_M2 -4.391 0.125 306.752 0.980
|Printer friendly Cite/link Email Feedback|
|Author:||Yin, Peng; O'Neal, Dennis L.|
|Date:||Jul 1, 2014|
|Previous Article:||Preliminary test and analysis of a stirling-engine-based residential trigeneration system at TRCA archetype sustainable house.|
|Next Article:||Characterizing airflow and power of VAV series fan-powered terminal units from component data--part II.|