Saving Energy With VFDs
May 5, 2006
The refrigeration load on a vapor compression system can be highly variable. For this reason, some means of adjusting the system's evaporator capacity to maintain temperature control of refrigerated spaces is essential. Refrigeration systems are commonly set up to control evaporator capacity by cycling the liquid refrigerant feed solenoid to maintain space temperature within a dead-band range around a desired setpoint. As the space temperature increases (an indicator of increasing load), additional evaporator liquid feed solenoids open in an attempt to lower the space temperature back to its setpoint. As the space temperature drops (an indicator of decreasing load), the liquid feed solenoid valves on operating evaporators are closed to prevent overcooling the space.
The simplest means of evaporator fan control is to operate the fans continuously. In the continuous fan operation strategy, capacity control is achieved by cycling the refrigerant feed to the evaporator, as outlined above. Although simple, this approach will maximize both the direct fan energy and the parasitic refrigeration load on the space attributable to fan motor operation. The most notable advantage of continuous fan operation is that constant air movement will minimize temperature stratification within the refrigerated space.
A variation on this theme is to duty cycle the fans when the solenoid remains closed for a given dwell time. An example of this control would be to de-energize the fan when the temperature requirements remain satisfied for 30 minutes. Because prolonged periods of fan inactivity can lead to temperature stratification, it is not uncommon to cycle fans on even when the space temperature is at setpoint for an extended period of time. The advantage of this method is that the electrical power and parasitic load associated with fan operation is shed during low load periods.
One other option is variable speed fan operation using variable-frequency drives on the fan motors. In this strategy, the fan speed will increase or decrease as the space temperature deviates above or below setpoint, respectively. Because reducing the fan speed results in a reduction in torque, the power required by the motor drops dramatically. For example, if the fan speed is reduced to 50 percent of its full speed, the required fan power will decrease by 87.5 percent.
How does the capacity of an evaporator vary with fan speed? The full-load capacity of an evaporator will depend primarily on the difference between the saturated evaporating temperature and the space setpoint temperature. Other factors that influence the full-load capacity of an evaporator include:
- Air volume flow rate (cfm).
- Superheat (i.e., direct expansion vs. overfeed).
- Subcooled liquid feed (mechanically pumped overfeed evaporators fitted with evaporator pressure regulators [EPRs]).
As the speed of an evaporator fan is reduced, the capacity of the evaporator decreases as well (although this reduction is not linear).
Part-Load Evaporator EfficiencyThe efficiency of the evaporator can be expressed in terms of the required brake horsepower (BHP) per ton of refrigeration load and varies as a function of the ratio of the load to the evaporator capacity. For continuous fixed speed fan control, the power requirement of the fan is the full-speed power, and matching the load is accomplished by cycling the liquid refrigerant feed solenoid. For aggressive fan cycling control, the refrigerant feed solenoid and the fan are often cycled together; however, this option is not recommended.
In the variable-speed case, the control strategy preferentially uses fan speed as the first stage of capacity control until the minimum speed is reached (usually at a frequency between 20 to 30 Hz). During reduced speed operations, the liquid feed solenoid can be cycled as well to better match the reduced load on the evaporator and its reduced need for cold liquid refrigerant. Once at minimum speed, further capacity reductions are achieved by cycling the liquid feed solenoid. Figure 1 shows the fan horsepower per ton of space load over a range of evaporator part-load conditions for all three fan control strategies.
It is clear from the figure that the variable-speed fan control has the best efficiency advantage at low part-load operating conditions, but the efficiency advantage decreases at higher part-load conditions. From a purely financial standpoint, the energy savings that results from the efficiency advantage of a variable-speed evaporator fan operating at part-load conditions needs to be sufficient to justify the capital cost premium associated with the drive controller. In other words, as with all VFD applications, part-load operation must be used enough hours to “pay” for the drive.
Other ConsiderationsFigure 1 only considers the influence of direct fan horsepower; however, fan power is a “parasitic” evaporator load, and any reduction in fan horsepower also will translate into a reduced load on the compressors. Keep in mind that a 5-hp fan power reduction is equivalent to a 1-ton reduction in refrigeration load. Because VFDs on evaporator fans offer the opportunity to significantly reduce the fan power, it is important to also incorporate the reduced parasitic load into a broader analysis that would include compressor and heat rejection systems.
Finally, keep in mind that inefficiencies are associated with the conversion of electrical energy (what we pay for) to mechanical energy (i.e., horsepower to turn the fan).
Motor and VFD Efficiencies. Estimating the electrical power requirement for a VFD-driven device requires both the motor efficiency over the expected range of part-load horsepower and the VFD efficiency over the expected range of speeds. A 1999 ASHRAE Journal article by Bernier and Bourret highlighted some of these motor/VFD issues.1 This information was further used by Chan to emphasize the inclusion of these inefficiencies when assessing VFDs.2 Because smaller horsepower motors are more common on evaporators, the estimation presented here assumes that the shape of the efficiency curve is the same but the nameplate efficiency is lower for the smaller motors. The VFD efficiency is handled the same way -- the efficiency curve is the identical shape but is scaled to full speed. NEMA requires full-load motor efficiencies of 85 to 90 percent for 1- to 5-hp premium efficient motors. Nominal VFD efficiencies are typically in the 94 to 96 percent range.
The application of the drive and motor efficiencies allows an estimate of the power consumption of the motor/drive combination. The full-load motor efficiency is assumed to be 88 percent, and the full-speed VFD efficiency is assumed to be 95 percent. Included in figure 2 are field data from a 15-hp for a VFD fan motor application. Although this data is for comparison only, it shows that the present analysis and assumptions reasonably predict the power of the fan motor and the VFD at a lower speed operation.
Parasitic Load Reduction and Compressor Efficiencies. To account for the parasitic fan motor loads, the fan's power consumption is converted to a load and added to the full-load compressor efficiency (hp/ton).
Overall Evaporator Part-Load Efficiency. The resultant fraction of full-load power as a function of evaporator part-load ratio is shown in figure 3 for a -20oF freezer space with seven evaporators and a saturated suction pressure of approximately 14.5 psia (0.4 in Hg vacuum), which corresponds to a combined efficiency of approximately 2.0 hp/ton. Clearly the VFD excels at low part-load conditions. At a part-load ratio of 50 percent, the VFD control requires 20 percent lower power (kW) for each ton of refrigeration when compared to the continuous fixed speed fan option. The VFD control retains a 9 percent efficiency advantage of the duty cycling at 50 percent as well.
Energy AnalysisNow that the analysis has been extended to include all of the inefficiencies and load reduction benefits of variable speed fan operation, the energy analysis allows estimates of operating cost savings.
For the previous freezer example, assume that those seven evaporators have a total capacity of 100 tons at the 8.5oF temperature difference (TD) at design conditions. The average load over the course of the year is 82.5 tons, with the assumed distribution as shown in figure 4.
Multiplying the kW/ton data from figure 3 to the load and hours data from figure 4 gives the total electrical energy consumption (kWh) associated with the compressor and the evaporator fan. Table 1 shows the yearly kWh per peak load (100 tons) for the three fan control methods discussed earlier.
Multiplying the savings from table 1 by the electricity rate and designed freezer load for a given location results in an estimate of yearly savings for the VFD applied to the evaporator fans.
Economic results for the application of VFDs for $0.05/kWh results in about $50/ton of design evaporator load when compared to continuous fixed speed (FS) fan control, and about $30/ton when compared to duty cycling (DC) fan control.
Beyond Energy SavingsVFDs offer many benefits when applied to evaporator fans. Probably the most appealing is the reduction in electrical energy usage; however, other ancillary benefits include fewer system transients, better space temperature control, reduced “wind chill” and noise, an inherently soft start and an improved power factor.
Note that the energy benefits associated with applying VFDs to evaporator fans depend strongly on the available evaporator capacity over the entire range of refrigeration loads. Systems with greater evaporator capacity (i.e., a larger surface area) will benefit more from a variable frequency drive than systems that are short of evaporator capacity
Further ReadingThe Northwest Energy Efficiency Alliance (NWEEA) has done a series of projects aimed at identifying the benefits and barriers associated with the application of VFDs to refrigeration evaporator fans. The Evaporator Fan VFD Initiative reports are available at www.nwalliance.org.
References1. Bernier, M.A., B. Bourret, “Pumping Energy And Variable Frequency Drives,” ASHRAE Journal, December 1999, Atlanta, GA, 2004.
2. Chan, T., “Beyond the Affinity Laws,” Engineered Systems, August 2004.