1、AN-04-3-2 Total Energy Consumption Model of Fan Subsystem Suitable for Continuous Commissioning Fu I in Wang Harunori Yoshida, Ph.D. Member ASHRAE Masato Miyata ABSTRACT This paper describes a newly developed total energy consumption model of a fan subsystem, which consists of fun, driveline, motor,
2、 and variable-speed drive. How to use this model for automated continuous commissioning is explained and verijied usingan experimental study. This model can accu- rately simulate the total energy consumption of a fan subsystem, which can be easily measured. Therefore, this model is useful for the au
3、tomated continuous commissioning of u fan subsystem during operation. The simulation results were verified using measured data in a real variable-air- volume system. The average diference between the simulated and measured datu is 5.1 %, which is accurate enoughfrom an automatedcontinuous commission
4、ingpoint of view to monitor the operation of a fan subsystem and detect faults during oper- ation. An experiment demonstrated that loose fan belts can be detected and the results are discussed as an example showing that the model can successfully detect this fault. INTRODUCTION Building commissionin
5、g is the process of ensuring that building systems are designed, installed, functionally tested, and capable of being operated and maintained to perform in conformity with the design intent (ASHRAE 1996). The awareness that commissioning is a viable method to help ensure buildings and their energy c
6、onservation measures (ECMs) meet design intent has been gradually growing since the 1980s. Some analyses of the data from the buildings partic- ipating in an energy conservation program revealed that many installed energy efiiciency measures were not performing as expected (BPA 1992). The main reaso
7、n is that the installed ECMs had not been properly commissioned. Building commissioning begins with the program phase and continues through the design phase, construction phase, acceptance phase, and post-acceptance phase (ASHRAE 1996). Post- acceptance phase commissioning is to continuously commis-
8、 sion the building systems to make them always run efficiently during their whole life cycle. For the purpose of automatically and continuously commissioning fan subsystems using simulation analysis during the operation phase, currently available fan simulation models were checked to determine wheth
9、er or not these models are suitable for continuous commissioning. Model validity checking showed that no fan model can give simula- tion results that match the experimental measured data quite well. Therefore, a new total energy consumption model of a fan subsystem was developed, which is suitable a
10、nd useful for continuous commissioning. This newly developed models accuracy and validity for continuous commissioning were verified using experiments. MODELS The fan models used by currently available simulation tools can only simulate the performance of a fan itself. There are no models to simulat
11、e the performance of other compo- nents in a fan subsystem, such as motor, inverter, etc. For example, SIMBAD can simulate a fans energy consumption using an empirical equation by inputting the real-time and maximum airflow rate and maximum energy consumption (CSTB 2001). HVACSIM+ is able to simulat
12、e a fans energy consumption and a fans pressure head using airflow rate and fan rotation speed by a series of equations fitted using manu- facturers data (Clark 1985). Fulin Wang is a Ph.D. candidate, Harunori Yoshida is a professor, and Masato Miyata is a masters student in the Department of Urban
13、and Environmental Engineering, Kyoto University, Kyoto, Japan. 02004 ASHRAE. 357 30% I 0.4 05 06 07 O8 o9 Cf 1800 rlm 8 1400 rlm A 1000 rlm Figure 1 Fan eficiency vs. dimensionless airflow rate. However, it is difficult to commission a fan using these models because they simulate the energy consumpt
14、ion of a fan itself and this energy consumption, which is termed fan shaft power, is difficult to measure. Especially during the operations phase, there is no building energy management system (BEMS) that measures a fans shaft power; but it is very easy to measure the total power consumed by the fan
15、 subsystem, which includes fan, dnveline, motor, and variable-speed drive (VSD), using a power meter set at the power input point such as the power switch. If the above-mentioned fan models are used to simulate the total power consumption of a fan subsystem, they will not give acceptable results. Fo
16、r instance, SIMBAD gives an average difference of 50%, and HVACSIM offers an average difference of 48%. Therefore, it is unreason- able to use these fan simulation models for continuous commissioning. Furthermore, the characteristics of total effi- ciency of a fan subsystem are different from that o
17、f the fan itself. A fans efficiency changes only according to dimension- less air-flow rate Cf The 7- Cfcurve shows a uniform shape at different fan rotation speeds, as shown in Figure 1, whereas the total efficiency of a fan subsystem changes according to not only Cf but also fan rotation speed, as
18、 shown in Figure 2. Therefore, new total efficiency models of fan subsystems are needed to express the unique characteristics of a fan subsystem. To address this problem, a total energy consump- tion model was proposed, which takes into account the perfor- mance of all the components in a fan subsys
19、tem, Le., fan, driveline, motor, and VSD inverter. The components and energy flow of a fan subsystem are shown in Figure 3. The total energy consumption model is defined by the following equa- tions. VAP E, = - qt 75% 100% AHU outlet 50% I damperopening E 45% : 35% 2 ._ 5 r 8 40% 6 30% =-. * 25% ; 2
20、0% /! - 2 15% 10% : , O 0.1 O2 03 04 05 06 07 Cf o 1800 r/m A 1350 rim x 900 r/m Figure2 Total eficiency of fan subsystem vs. Dimensionless airflow rate. Air, VAP AA Figure 3 Components and energyflow of a fan subsystem. (4) q, = mo+mlL+m2L 2 +m3L 3 +m4L 4 (5) 2.3 4 q. = io + ilL + i2L + zL + i4L c
21、=P u2 APD4 (9) V e-, -ND Equation 1 is used to calculate the total energy consump- tion of a fan subsystem using airflow rate, pressure head, and total efficiency of the fan subsystem. Total efficiency of a fan subsystem is calculated using Equation 2 by multiplying the efficiency of all the compone
22、nts ofthe fan subsystem, i.e., fan, driveline, motor, and inverter. Equation 3 is used to calculate 358 ASHRAE Transactions: Symposia Figure 4 Comparison of measured and calculated fan rotation speed. fan efficiency using a fourth-order function of dimensionless airflow rate. This equation is taken
23、from HVACSIM+ (Clark 1985). Motor and inverter efficiency is calculated using Equa- tions 4 and 5, respectively, which are fourth-order functions of load factor. Load factor is calculated using Equation 6, which is the rate of fan shaft power to rated shaft power. As discussed below, Equation 7 is u
24、sed to calculate the dimensionless airflow rate using dimensionless flow resis- tance. Dimensionless flow resistance is calculated using Equation 8, instead of the method used in HVACSIM+, which uses airflow rate and fan rotation speed as shown in Equation 9. Fan rotation speed is estimated using Eq
25、uation 10 from rated rotation speed and inverter output frequency. Fan Efficiency Model Equation 3 shows a fan efficiency model used in HVAC- SIM+. This fan efficiency model simulates a fans efficiency using dimensionless airflow rate C which is calculated using airflow rate and fan rotation speed.
26、However, for a gear driven or cog-belt driven fan subsystem, the fan rotation speed can be calculated using inverter output frequency and rated rotation speed, as shown in Equation 1 O, because no slippage occurs in gear drivelines and cog-belt drivelines. Whereas for a v-belt or band belt driven fa
27、n subsystem, the belt slippage makes it inac- curate to calculate the transient fan rotation speed using inverter output frequency and rated rotation speed. Further- more, most BEMS are not installed with a sensor to measure the fan rotation speed. Therefore, for the purpose of continu- ous commissi
28、oning, other models must be used. One alternative method is to simulate the fan efficiency using airflow rate and fan pressure head. For this purpose, a new dimensionless variable was proposed, dimensionless airflow resistance coefficient C, as shown in Equation 8. Using C, and Equation 7 to calcula
29、te C fan efficiency, can be calculated using Cfand Equation 3. Another method is to estimate fan rotation speed using inverter output frequency, as shown in Equation 10. The accu- racy that this estimation could achieve was checked by comparing the measured rotation speeds with estimated ones. The a
30、verage difference between the measured and estimated rotation speed is 2.2% and the maximum difference is 9.0%, as shown in Figure 4, when inverter output frequency gradu- ally changed from 20% to 100% of rated frequency, 50 Hz. The accuracies of these three fan efficiency simulation meth- ods were
31、compared to determine which one is most suitable for continuous commissioning. The input variables to the three simulation methods are as follows. Airflow rate and head pressure Airflow rate and measured fan rotation speed Airflow rate and calculated fan rotation speed The average differences betwee
32、n the measured total effi- ciency of fan subsystem and simulated total efficiency given by these three methods are 5.1%, 6.1%, and 6.5%, respec- tively. The simulation model using airflow rate and head pres- sure gave the most accurate simulation results. The simulation accuracy using measured fan r
33、otation speed is 0.4% higher than using calculated fan rotation speed. So it is acceptable to simulate a fans efficiency using calculated fan rotation speeds instead of measured ones when head pressure is unavailable. Driveline Efficiency The newly installed drivelines efficiency can be used as a pe
34、rformance target value during the continuous commis- sioning phase because the driveline efficiency changes due to aging. So as a drivelines efficiency decreases because of aging, the real total efficiency of a fan subsystem will be lower than the total efficiency simulated using the newly installed
35、 drivelines efficiency. Therefore, a newly installed drivelines efficiency can be used in the fan subsystem simulation to help detect the fault of belt aging and is useful for continuous commissioning. The recommended efficiency value for a newly installed V-belt is 95% and for a newly installed cog
36、ged or synchronous belt efficiency it is 98% (OIT 2000). Motor Efficiency Motor efficiency changes according to two variables, electric frequency and load factor, as shown in Figure 5. Load factor is the percentage of transient motor power output to rated power output. If fan belt efficiency is assu
37、med to be constant at various loads and rated motor output is equal to rated fan power input, the load factor equals the transient fan power input divided by rated fan power input. For VSD-driven motors in a VAV system, the electric frequency input to the motor is not an independent variable. It is
38、determined according to demand airflow rate and pressure, which determine the load on the motor. Therefore, required motor input frequencies are related to load factors. The rela- tionship between load factor L and required frequency F can ASHRAE Transactions: Symposia 359 Figure 5 Motor ejciency vs
39、. load factor and frequency. be derived, as shown in the Equations 1 1 to 13. From Equation 13 it can be found that the required electric frequency has one definite value corresponding to one load factor value. There- fore, load factor is the single independent variable that can influence the motor
40、efficiency. So for a VSD-driven motor in a VAV system, the relationship of motor efficiency and frequency and load factor is not a three-dimensional surface, but three-dimensional curves, as shown in Figure 6. In order to simulate the motor efficiency accurately, Equation 4 needs to be fitted using
41、the data with the relation- ship of vrn - L - (F / Fr)3, as shown by curve A in Figure 6. However, most motor manufacturers only offer data on their motors efficiency at the conditions of rated frequency and various load factors, as shown by curve B in Figure 6. There- fore, these data have to be us
42、ed to fit the motor efficiency model. Of course it is not accurate to simulate the motor effi- ciency in a VAV system using such fitted model. So motor manufacturers should be urged to offer motor efficiency data with the relationship to load factor and frequency at the condi- tions of vrn - L - (F
43、/ Fr)3. AP = S$ (12) Inverter Efficiency Inverters change the electric frequency to make motors and fans rotation speed variable. During frequency modula- tion, inverters consume some energy, which becomes heat and discharges to the local environment. Inverter manufacturers literature shows the rela
44、tionship between heat-generating Figure 6 VAV system motor ejciency vs. load factor and frequency. 110% 100% 93 90% B o .g 80% lo 8 70% $ 60% 50% 40% E m y xverter ADacilv 0.75 kW or less Inverter capacity 1 5 to 7 5 kW K Inverter capacity 11 kW or more I 0% 20% 40% 60% 80% 100% 120% Load Factor Fig
45、ure 7 Inverter heat generating vs. load factor: losses and load factors, as shown in Figure 7 (Hitachi 2002). Using these heat-generating rates, the inverter efficiency can be calculated as a function of changing load factor. Using these data of inverter efficiency vs. load factor, the inverter effi
46、- ciency model can be fitted. The fitted inverter efficiency model can be used to simulate the inverters efficiency during the operations phase to realize continuous commissioning of a fan subsystem. EXPERIMENTS In order to verify the simulation accuracy of this total energy consumption model of fan
47、 subsystem and its validity for continuous commissioning, experiments were conducted in August 2002 on a real VAV system in an office building located in Tokyo. In the first experiment, normal fan subsystem operation data were measured to verify the total energy consumption models accuracy. The seco
48、nd experi- 360 ASHRAE Transactions: Symposia Air Velocity 0 uu Probe -. Humidifier /4 I Filter Damper Return Air Fan Chilled /t +Chilled Water Inlet Water Outlet 100% Inverier 75% I 50% output T- b AHU 100% 100% Outlet Damper Opening I b 10:45 11:OO 11:15 11:30 11:45 1200 1325 13:40 1355 14:lO 1425
49、14:40 Figure 9 Experiment step and parameters. 50% Figure 8 Measurement points and AHU conJiguration. 40% x ment checked whether this total energy consumption model could detect the fault of loose belts. The total electric power fan in an AHU were measured. The measurement points and the configuration of the AHU are shown in Figure 8. 8 30% consumption, air volume, and head pressure for the supply air 20% -Measured total efficiency of fan subsystem -Simulated total efficiency of fan subsystem 10% Normal Fan Subsystem In order to obtain the fan performance d
