An advanced control system can provide significant energy savings while also optimizing the operating behavior of chillers with inverter-driven compressors.

Chillers with variable-speed, inverter-driven compressors can provide energy savings compared to conventional on/off compressors. However, the reductions in cooling capacity often associated with variable-speed compressors can have a significant impact on chiller control, particularly with regard to compressor and water limits and overall unit stability. Managing a chiller with a variable cooling capacity therefore requires a high level of communication between all of the controlled devices and the control system to allow the operator to easily identify and correct any problems.

A new chiller control system has been developed that is based on an enhanced communication technology. By improving control of inverter-driven compressors, electronic expansion valves and condenser fans, this system can provide energy savings while optimizing chiller performance in a range of operating conditions.

Figure 1. As the water temperature started to rise, the PID inverter control reacted to maintain the water setpoint of 56.03°F (13.35°C). The control algorithm increased the inverter frequency from 45 to 65 Hz, and both the valve and condenser control followed the variation in the load.

Test Conditions

To test the capabilities of the control system, the device was installed on a water-to-air heat pump chiller with a nominal cooling capacity of 16 kW (see sidebar: Test Chiller Description). A water tank was installed to control the load power at the evaporator. The water flow at the evaporator was controlled through a simple water circuit with a bypass valve and was measured with a flow meter, while the water temperature was controlled by four heaters with solid state relays. This setup allowed all water conditions and load powers for the unit to be simulated.

The chiller was installed in a climate chamber that was capable of reaching ambient temperatures between 59 and 95°F (15 and 35°C). All of the other ambient conditions were simulated using a calculated condensation profile. All calculations were based on a unit with fans operating at a constant maximum speed (maximum efficiency) so that a precise condensing temperature existed for each ambient temperature.

Eight temperature probes (Pt1000 sensors) and eight pressure probes (ratiometric pressure transducers) were installed at each component inlet and outlet to monitor both the pressure drops and temperature conditions. This information was used to determine the complete refrigerant cycle for the pressure-enthalpy (P-h) diagram. Temperature probes also were installed to measure water inlet and outlet temperatures and condenser air temperatures.

A water flow meter, a refrigerant mass flow meter and a wattmeter that measured compressor power consumption (including the inverter power supply) were installed to assess the chiller’s power balance. Although the power consumption of the heater easily could have been calculated for water temperature control based on the duty cycle of two identical pairs of 10 kW heaters, a voltage measurement device and three computed tomography (CT) probes were installed to obtain actual power absorption measurements.

The data acquisition software was developed on a PC-Windows platform, based on National Instruments’ LabVIEW 6i software, and communicated all of the acquired and calculated data (such as saturated temperatures) to the main control program for chiller and water control. All data were acquired with a cycle time of 0.5 sec, and the recordings were saved to a Microsoft Excel “history” file every 5 sec.

Table 1. The improved performance of inverter control at maximum efficiency (with the lowest condensing temperature at each ambient temperature) provided energy savings in all test situations

Control System

The chiller’s control algorithm was composed of four different parts:
  • Compressor control (with or without inverter) based on the water temperature.
  • Electronic expansion valve control based on the superheat value.
  • Condenser fan control based on the saturated condensing temperature.
  • Prevention and alarm procedures.

Compressor Control. A control system that could work with or without the inverter was developed so that comparative tests could be performed. Both types of control followed the water temperatures (inlet or outlet); however, control with the inverter required a proportional-integral-derivative (PID) algorithm for the frequency. The algorithm was based on the inlet (default) or outlet water temperature, and the inverter frequency was increased or decreased as needed to reach and maintain a water setpoint.

On/off duty also could be simulated by using the inverter at a fixed frequency. The compressor control had a simple water temperature setpoint with a differential that had to be reached to restart the compressor. Default values for the tests set the water inlet temperature at 54°F (12°C) with a 3.6°F (2°C) differential.

A control algorithm with PID parameters was designed to optimize the water temperature stability and the fastest-duty temperature achievement. When the unit was started, the initial frequency was calculated proportionally based on the difference between the actual water temperature and the setpoint. The compressor was stopped when the inverter frequency went below the minimum value (default 25 Hz).

Electronic Valve Control. The electronic valve was driven by a standard control device developed to move a stepper motor valve according to the superheat setpoint, which was measured by two probes installed at the evaporator outlet. The first probe measured the evaporating pressure, while the second one measured the superheated gas temperature. The control device was a stand-alone instrument that communicated the input probe values, valve position, control parameters and status with the unit main control (PC).

The main superheat control used in the test had a PID algorithm that worked with just the superheat and setpoint. However, many different situations exist in actual applications in which the chiller’s operating conditions (low superheat, high condensing or evaporating pressures, low evaporating pressure, etc.) are crucial. In these cases, it is important to have a control algorithm that can change the standard behavior of the electronic valve to prevent the unit from stopping due to high or low pressures or from working in critical conditions for the compressor. In many cases, the superheat setpoint or valve control of the advanced control system can be changed based on the probe values to decrease or increase refrigerant flow and cooling efficiency automatically and keep the unit working even in difficult conditions, such as under a heavy load or during high outside temperatures.

The valve control algorithm provides protection functions for all of the chiller’s warning situations (low superheat, high condensing temperature, low or high evaporating temperature) that can be avoided by adjusting the valve position. For the test, four different limits were used:
  • Maximum condensing temperature: 138°F (59°C).
  • Maximum evaporating temperature: 64°F (18°C).
  • Minimum evaporating temperature: 27°F (-3°C).
  • Minimum superheat value: 37°F (3°C).

Condenser Control. The condenser control used for the test had the same PID function as the compressor. The fan speed was controlled based on the condensing temperature setpoint. The main difference between the parameters was the correlation between the input and output signal.

The water temperature decreased with increasing compressor frequency (any other parameter should be constant) with a linear correlation. The increase in the compressor frequency was constant.

The condensing temperature was correlated with the fan speed, which had a cubic correlation with the voltage signal from the control. For this reason, the increase in fan voltage had a quadratic correlation with the voltage itself. This correlation was calculated using the last value of the fan voltage.

Prevention and Alarm Control. The prevention algorithm had a significant influence on the behavior of the chiller because it managed the same critical situations listed for the valve control but controlled the compressor frequency directly. Four prevention and alarm limits were established:
  • Condensing temperature: prevention, 142°F (61°C); alarm, 149°F (65°C).
  • Evaporating temperature: prevention, 23°F (-5°C); alarm, 18°F (-8°C).
  • Compressor discharge temperature: prevention, 248°F (120°C); alarm, 257°F (125°C).
  • Water outlet temperature: prevention, 37°F (3°C); alarm, 36°F (2°C).

In all of the prevention conditions, the inverter frequency was decreased with a defined step as soon as the prevention limit was reached, and then proportionally depending on the difference between the actual temperature or pressure values and the prevention limits.

For actual applications, a calculation model is being developed that will enable the control system to evaluate on a case-by-case basis whether warning conditions have to be managed by changing the valve control or decreasing the inverter frequency (and cooling capacity). Chiller operators and facility managers will likely be able to set the behavior of the unit in a “maximum cooling capacity” or “maximum efficiency” (coefficient of performance [COP]) mode.

Figure 2. The unit performance increased significantly, which lowered the load power with a maximum COP of 4.55 (at around 50 percent of load power).

Water Control

The load power control for the test was completely separate from the chiller control to avoid any kind of mutual influence. However, both control systems read and used the same temperatures from the data acquisition program.

The temperature control in the advanced control system works with pulse width modulation (PWM). The chiller operator or facility manager can select direct control of the load power (in kW) with a feedback measurement of the actual total power consumption of the heaters, or set a water temperature in the tank (which corresponds to the inlet water temperature of the chiller unit) with the same PID algorithm as the compressor. The control algorithm reads the water temperature and modulates the PWM signal from 0 to 100 percent (0 to 20 kW) to reach the setpoint. The water temperature or power profiles can also be automatically set to simulate different load conditions.

Figure 1 shows the control system’s front panel as well as the control parameters and a sample of the test chiller control in the graphs on the right. The colored lines are the input variables and setpoints, water temperature, superheat and condensing temperature, while the white lines are the output variables, inverter frequency, valve position and fan voltage.

This example shows the control reactions that occurred during a variation in the load power. As the water temperature started to rise, the PID inverter control reacted to maintain the water setpoint of 56.03°F (13.35°C). The control algorithm increased the inverter frequency from 45 to 65 Hz, and both the valve and condenser control followed the variation in the load. Figure 1 also shows the changes that occurred in the water and superheat setpoint.

Figure 3. A comparative test at 68°F (20°C) ambient temperature and 10.3 kW load power (around 70 percent of nominal load power) revealed that the power consumption was lower for the inverter-based system.

Inverter Control Characterization

One of the primary advantages to using an inverter-driven compressor is the relationship between the condensing pressure and power consumption. Compared to on/off duty, the inverter control modulates the compressor frequency at lower values. As a result, the system generates a lower condenser load -- and, consequently, a lower condensing temperature -- at a constant ambient temperature.

Tests performed on a system with maximum efficiency control and a system with constant condensing pressure control at the same ambient temperature (68°F [20°C]) and load conditions (40 to 100 percent) showed that the inverter frequency decreased as the load decreased in both cases, and the condensing temperature and related power consumption of both units were also lowered.

As figure 2 shows, the unit performance of the chiller with the advanced control system increased significantly, which lowered the load power with a maximum COP of 4.55 (at around 50 percent of load power). The light green line shows the maximum performance COP curve compared with the dark green line, which represents the same inverter performance without the contribution of the lower condensing pressure control.

The test also revealed that the unit COP decreased around both the maximum (75 Hz) and minimum (25 Hz) frequency due to the inverter power consumption and compressor design. A frequency of 75 Hz is encountered in actual operations only during about 1 to 2 percent of the total running time and is therefore unlikely to cause problems in overall chiller performance. However, controlling the minimum inverter frequency to avoid the compressor operating with a low COP can provide energy savings. For this reason, the minimum inverter frequency of the advanced control system is calculated by the control algorithm on a case-by-case basis and is typically set between 30 and 35 Hz, depending on the outside temperature. Working with an on/off modulating frequency stopped at the minimum value can be more efficient than working at constant low frequency values.

The primary goal of the comparative tests was to measure the real influence of on/off duty on the power consumption of the chiller unit. The test specifications started with on/off duty at a fixed 75 Hz inverter frequency regulating a variable heat load with a constant water setpoint (54°F [12°C], 3.6°F [2°C] differential). Power consumption and temperatures were measured and average values were calculated to obtain water setpoint values for the inverter modulating frequency test. These parameters allowed the mean power consumption to be measured with the same water temperatures and load in both situations. Figure 3 shows the mean values of the chiller’s cooling efficiency (water temperature with a constant load power) and power consumption in both tests.

Some of the comparative tests are summarized in table 1. Both types of inverter control are listed and compared with on/off duty to measure the exact increase in power consumption due to the inefficiency of on/off control and the better performance of inverter control at maximum efficiency (with the lowest condensing temperature at each ambient temperature).

As previously mentioned, the test unit experienced high power consumption at low inverter frequencies (below 35 Hz) due to the low COP of the compressor-inverter system when running at a low speed. Table 1 (line 4) shows a typical condition in which on/off duty is 7 percent more efficient than inverter duty and in which the inverter control works better by stopping the compressor at 25 Hz and starting it at around 50 Hz, with a modulation period in between.

Improved Energy Savings and Control

Comparative tests have shown that chillers based on inverter-driven compressors consume up to 15 percent less energy than chillers based on conventional on/off compressors, depending on the load power. Even greater efficiencies can be achieved by lowering the condensing pressure -- inverter-driven compressors can consume up to 30 percent less power at medium range inverter frequencies. However, an advanced control system is needed to optimize control of a chiller with an inverter-driven compressor.

The direct communication provided by the new control system between the inverter, valve and condenser allows all of these components to operate in harmony, with a high level of water stability and effective reactions to all variations in load and ambient conditions.

Sidebar: Test Chiller Description

  • Unit type: water-to-air heat pump chiller
  • Refrigerant: R407 C
  • Nominal cooling capacity: 16 kW (Tcond = 104°F [40°C], Tevap = 39°F [4°C], 75 Hz, Sh = 50°F [10°C], Sbc = 50°F [10°C])
  • Compressor type: inverter-driven scroll compressor, 5 hp
  • Inverter type: multi-purpose three-phase inverter, 380 to 460 V ±10%, 50/60 Hz ±5%
  • Condenser type (chiller): finned coil
  • Evaporator type (chiller): plate
  • Primary expansion device: stepper motor electronic expansion valve
  • Secondary expansion device: thermostatic valve