ESL-IC-5-1- Development of Power-head Based Fan Airflow Station Gang ang Research associate University of Nebraska, Lincoln Mingsheng Liu Professor University of Nebraska, Lincoln Abstract Fan airflow measurement is critical for heating ventilation and air conditioning (VAC) system operation, control, and fault detection. It has been a challenge for VAC professional. Since curves under a given speed can relate airflow to head or power, both head based and power based airflow stations have been developed and implemented in many air handling units (AUs). The airflow station obtains the airflow based on the measured curve, speed and head or power. owever, it is hard to obtain accurate curves since airflow measurement is still required to determine curves. On the other hand, the airflow can be obtained by measured power and head with efficiency curve. Since the efficiency varies slightly when the VAC system airflow changes, the power-head based airflow station mainly depends on head and power measurement. The paper presents the theoretical models and the experiments of the power-head based airflow station. Introduction In a variable air volume (VAV) air handling unit system, the supply airflow varies with building load using variable frequency drive (VFD) to modulate supply speed to maintain a static pressure set point. To maintain the appropriate positive building pressure, the return airflow must be properly tracked with the supply airflow. The return airflow should be slightly less than the supply airflow. The difference between the supply and return airflow depends on building exhaust airflow and envelope tightness. One of tracking measures is to modulate the return speed to maintain the difference between the supply airflow and return airflow. The airflow is measured using flow stations in the main supply duct and in the main return duct [1]. In a laboratory exhaust system, the exhaust airflow varies with the fume hood sash position. Therefore the laboratory exhaust system is a VAV system. On the other hand, the stack exit velocity should be controlled for adequate dilution. Therefore multi-stack system is developed for the requirement of exit velocity. The exit velocity can be controlled in a required range by using different stack combination based on the measured actual building exhaust airflow []. The airflow measurement plays an important role in the VAC system operation. owever, the airflow station requires a straight duct for -1 duct diameters upstream and duct diameters downstream. There are very few systems that have such duct run in the main ducts. In order to obtain accurate airflow, Liu developed a airflow station to obtain the airflow using the speed and head based on the regressed curve [1]. The experiment was conducted to verify the station and an excellent agreement between the model and the experimental values was found []. The accuracy of the airflow station depends on the accuracy of the curve. Since the installation configuration is different with the testing condition, the actual curve is 1 Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5
ESL-IC-5-1- different with the manufacturer s curve. An in-site curve measurement was developed for VAV AU System []. involved. The paper presents the basic theory, experiment and results of the power-head based airflow station. The airflow station works very well in the high airflow range where the head is sensitive to the airflow. owever, the head varies slightly in the small airflow range. Therefore, it is difficult to obtain the airflow using the head in the smaller airflow range. Unfortunately, the may works in this airflow range under partial loads. On the other hand, in most of airflow range, the power curve varies exquisitely. ang and Liu developed the VFD airflow station to obtain the airflow using the power and speed based on the power curve. Both the airflow station and the VFD airflow station depend on measured head curve or power curve and measured speed. First, the airflow measurement is required to obtain the curve. As we mentioned previously, it is difficult to obtain accurate airflow in systems. It may result in an inaccurate curves. Secondly, the measured airflow is proportional to square of the measure speed using the head curve or cube of the measured speed using power curve. The accuracy is mainly related to the measured speed. Actually the measured speed is assumed to equal the synchronous speed, which is proportional to the VFD frequency. Theoretically it is not true. The difference between the synchronous speed and speed depends on the load. The less load results in the less speed differential. Therefore, it is hard to accurate curve and speed. Finally the curve based airflow has accurate problems. Theory Figure 1 shows variable speed connection schematic. VFD is normally installed on the to adjust the speed by modulating frequency. Typically the power input can be obtained from VFD output terminal. Both and have the mechanical energy loss, such as heat generation loss and friction loss. So the output mechanical energy can be expressed as: = = (1) Power input Fan Figure 1: mechanical connection Fan* Since the input power can be provided by VFD and the head is easy to be measured, the airflow can be obtained using this relationship if the and efficiency is known. = = () Based on the theory, the efficiency is the function of the power. On the other hand, the efficiency does not change much around design working point when the airflow changes. This gives us an opportunity to obtain airflow using the measured power and head directly without the speed and curve η = f ( ) () 1 Typically the efficiency can be considered as a constant when the power changes from 5% to fully load. Figure gives the efficiency Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5
ESL-IC-5-1- curve with the power provided by manufacture. 1 same ratio / no matter what the speed is. In other words, the efficiency is the function of /. Motor efficiency(%) 8 5 1 15 5 Motor power(k) η = ( f ) () For any airflow under a given speed, the head can be obtained from the head curve and the ratio / can be easily obtained. Then the Figure : Motor efficiency with power Typically the head and power curve under a given speed are provided by manufacture. Then the efficiency curve can be easily achieved from the head and the power curve. Figure shows the head curve and efficiency curve with the airflow under given speed for the backward centrifugal. Fan head(pa) 15 1 Fan head 1 8 9 5 1 15 Fan airflow (m/s) Figure : head and efficiency with airflow under a given speed (%) efficiency curve can be redrawn with the ratio / in figure. (%) 1 9 8 7 5 1 5 1 15 5 dimensionless factor Figure : efficiency versus Substitute Eqs. () and () into Eq. (1). / f ( ) f1( ) = (5) Figure also shows that efficiency is the function of airflow and speed. In order to eliminate the speed, the law is applied. It can be proved that the has a same efficient under a Therefore, the airflow can be obtained from the measured power and the measured head. The actual efficiency and efficiency Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5
ESL-IC-5-1- will change based on the system operation. Normally in the VAV system, the speed is modulated to maintain the duct static pressure at its set point when airflow changes. The curve control curve depends on the chosen design head and the static pressure set point. If the design working point is chosen to have the high efficiency and the static pressure set point is 1% and 5% of the design head respectively, the system control curves are shown in figure 5. Fan head(pa) 15 1 9 5% Fan head Design point System curve setpiont of 1% 5 1 15 Fan airflow (m/s) Figure 5: System control curve with different static ratio pressure set points 1 Since the efficiency is a function of / 8 (%), these system curves are redrawn in Figure with coordinates of airflow and ratio of /. The efficiency is also drawn in the efficiency changes from 75.% to 7.7% (from B to B1) for 1% setpoint and from 75.% to 7% (from B to B) for 5% of setpoint. It can be seen that the efficiency is not sensitive to the airflow change. The efficiency slightly deceases as the airflow decreases. (%) 1 11 9 A Deisgn point 1 8 B B1 9 B 7 8 7 5 A1 A 5 setpoint of 1% setpoint of 5% system curves 1 1 5 1 15 5 Resistance factor Figure : versus with airflow Motor efficiency(%) efficiency 1 1 8 Design point 1 1% setpoint of 5% 8 pow er under under airflow s 5 1 15 5 Motor power output(k) airflow(m/s) airflow(m/s) same chart. The ratio / under different Figure 7: Motor efficiency versus airflow airflow can be obtained by following different system control curves and then the efficiency can be Fan head under different airflows can be obtained based on the different system control curves, then the obtained by the ratio /. ith a setpoint of ratio / can be used to calculate the 1% of the design head, the ratio / changes from 1 to 15 (from A to A1) when airflow decreases from the design airflow to the half. ith a setpoint of 5%, the ratio changes from 1 to (from A to A). As results, the efficiency, and finally the power or power input can be obtained. On the other hand, the efficiency also can be expressed as the function of the input power instead of the output. Figure 7 shows the required power with the airflow change and the efficiency with the Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5
ESL-IC-5-1- power. The required power decreases from 15.7k to.5k with a setpoint of 5% of the design head and from 15.7 k to. k when the airflow changes form design value to the half. It can be seen that the efficiency is also not sensitive to the airflow change and slight deceases as the airflow decreases. In reality, the measured static pressure differential is not the head. The equivalent efficiency should replace the efficiency. The measured static pressure is the head minus the dynamics pressure and the pressure loss. So the measured static pressure differential can be expressed: Δ = S () The equivalent efficiency is expressed as: Δ ( η = = S ) / η = η (1 S ) = f ( ) (1 S ) (7) Replace the head by the measured pressure differential. η = ( Δ + S f ) (1 S ) Δ + S ( Δ S Δ = f + S) (1 ) = f( ) (8) Δ + S The factor S depends on the pressure sensor location. For a fixed location, S is constant. The equivalent efficiency is still determined by the ratio Δ /. Therefore, all equation can be used for the measured pressure differential and the equivalent efficiency. The airflow can be obtained by the measured power and measured pressured differential with the efficiency and the equivalent efficiency. f1( ) f ( ) = (9) Experiments In order to develop the power-head based airflow station, both the efficiency and efficiency should be known. The efficiency can be obtained from manufacture and the efficiency need be evaluated by experiments. The experiments were conducted on a laboratory fume hood exhaust system to validate the theory. The exhaust system has a with VFD. The VFD modulates the speed to maintain the duct static pressure at 1. inch of water. The system airflow changes from 5.m /s to.7m /s and the speed changes from % to 7% of the design speed when all fume hood sashes change from full closed position to full open position. The two stacks were designed in the system to ensure the required exit velocity all the time. Two stack are used when the airflow is higher than 5.m /s and one stack is used when the airflow is lower than 5.m /s. Since the system has a 15 meter long main duct without any branches, the airflow is measured based on the pressure drop through the main duct and then the stack damper is controlled by the measured airflow. During the measurement the airflow was obtained by the measured duct pressure drop and the power were obtained from VDF. Meanwhile, the head was measured using the pressure sensor. All the measured parameters are recorded in computer with a time interval. The measurements were conducted under three different speeds, 85%, 75% and 5% of design speed. Under each speed, the airflow was adjusted by changing fume hood sashes from full open position to full closed position. Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5 5
ESL-IC-5-1- Figure 8 shows the average head via the airflow under 85%, 75% and 5% of design speed when the sash positions were changed from closed to full open position. Figure 9 shows the power via the airflow under same conditions. Fan head (Pa) 1 9 8 7 VFD-85% 5 VFD-75% VFD-5% 1 8 Fan airflow (m /s) Figure 8: Fan head under different speeds f ( ) (%) Δ f ( 1 = ) Δ.87 = (11). motot 9 8 7 5 1 5 1 15 5 ratio Figure 1: Motor efficiency curve versus dimensionless factor Δ / Figure 1 shows the efficiency curve verse Motor power (k) 1 8 Output-85% Output-75% Output-5% dimensionless factor Δ / based on the measured data in figure 8 and 9. The efficiency curve can be regressed using linear expression with the dimensionless factor. Δ η =.9.1 (1) 8 Fan airflow (m /s) Figure 9: Motor power under different speeds It is found that the efficiencies in the actual power range can be regressed as a function of the power.. η = f 1( ) =. 87 (1) The equivalent efficiency curve can be obtained introducing Eq.(1) into Eq.(9), Finally the airflow can be expressed as the function of the measured power and head.. (.9.1 ) = (1) Conclusions The power-head based airflow station theory has been deduced that the airflow can be obtained using the power and head. A power-head based airflow station was developed in an existing exhaust system. The airflow station mainly depends on the power and head measurement, so the Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5
ESL-IC-5-1- accurate airflow can be obtained. Nomenclature - head, Pa; - airflow, m /s η - power, kw; - efficiency References 1. Liu M.,. Variable Speed Drive Volumetric Tracking (VSDVT) for Airflow Control in Variable Air Volume (VAV) Systems. Journal of Solar Energy Engineering,, Vol. 15, pp. 18-.. ang, G., Cui, Y., Yuill, D., and Liu, M.,. Development of Multi-Stack Exhaust Systems for Laboratory Buildings, Proceeding of ASME Solar Energy Conference, Reno, NV,.. Yuill D. P., Redmann N. K. and Liu M.,. Development of Fan Airflow station for Airflow Control in VAV systems. Proceedings of ASME Solar Energy Conference, ISEC, Big Island, awaii,. Liu M., Liu G., Joo I., Song L. and ang G. Development of In-site Fan Curve measurement for VAV AU systems. Proceedings of ASME Solar Energy Conference 7 Proceedings of the Fifth International Conference for Enhanced Building Operations, Pittsburgh, Pennsylvania, October 11-1, 5