Superheat

There are two basic schemes for sensing superheat. True superheat is a pressure-temperature relationship, specific to each refrigerant.

When electronically derived, pressure-temperature (P-T) superheat requires the use of a pressure transducer, a temperature sensor, and a pressure-temperature table or equation. A simpler but less accurate measure of superheat is the two-temperature method.

In the two-temperature method, the temperature is sensed at the inlet and at the outlet of the evaporator. The difference in temperatures is then assumed to be superheat.

Refrigerants or blends with temperature glides may affect two-temperature superheat control. Ordinarily, superheat setpoints must be higher to overcome the effects of glide.

An advantage to two-temperature superheat is cost; pressure transducers are far more expensive than thermistors.

Additionally, two-temperature superheat works with any refrigerant without reprogramming. The temperature difference between the two sensors will indicate superheat no matter what the pressure-temperature relationship of the refrigerant.

The main disadvantage of the two-temperature method is the uncertainty that the inlet sensor is located properly. For the two-temperature superheat method to be accurate, the inlet sensor must be located in a position that has saturated refrigerant present at all times. Failure to find, or use, the proper location can lead to poor control or compressor damage.

Because of these factors, it is unlikely that retrofitting EEVs with two-temperature control to existing systems in the field will ever become practical.

True pressure-temperature superheat is by far the most accurate means of EEV control. As the cost of accurate pressure transducers falls, the P-T-based controller will become much more common. Although these types of controllers require microprocessors with more memory, the cost of the necessary computer technology drops almost daily.

Within a few years EEVs may become economically competitive with their mechanical counterparts, and almost as easy to retrofit or install.

Once the superheat is determined, it’s time to create the software to control the valve.

Software

The valves (with their motors and wiring) and controllers (with their transistors and microprocessors) are grouped together as “hardware.”

To make the hardware perform a function, a set of instructions must be given to the microprocessor. This set of instructions is called “software.” Certain “routines” must be incorporated to make valve control possible.

Most step motor valves are designed without internal intelligence or feedback; that is, they move only in response to controller signals. The valves maintain their position when no signals are received and valve position is stored in controller memory.

When the valve is given a signal to change position, the controller keeps track of the change; however, the controller does not directly “know” whether the valve has changed position.

To make this form of control effective, two control routines must be implemented: initialization and feedback loops.

Initialization

Initialization occurs when the valves are powered up for the first time, and sometimes when a large change to the system is made (e.g., closing for defrost). When the controller-valve combination is first powered together, the control does not know the valve position.

To initialize, the controller sends out a stream of closing steps greater than the total number of steps in the valve stroke. This will ensure that the valve is closed. This closed position becomes the “0” (zero) position of the valve used in all subsequent controller calculations.

This series of extra steps is called “overdriving,” and the valves have been designed to accept this without damage. The actual number of overdriving steps depends upon the manufacturer and model of the valve used.

It is vitally important that the controller be configured for the valve being controlled; failure to do so may damage the compressor. If the controller is configured to control a valve having 50 steps and the valve actually has 100 steps of stroke, then the controller may not fully close the valve, leading to floodback.

Once the valve is fully closed and the controller knows the “0” valve position, the algorithm may be implemented with the aid of a feedback loop.

Feedback loops

Feedback occurs when the result of a process is sensed and the sensory information is used to modify the process. In simpler terms, when the controller opens the EEV too much, causing overcooling, the temperature sensor “feeds back” that information, and the controller closes the valve.

If a control algorithm were written with only references to the absolute number of steps open, then changes in head pressure, liquid temperature, etc., would not be taken into account and control would be poor.

Instead, sensors are used to ascertain the effect of valve position on temperature and the position is changed to bring the sensed temperature closer to the setpoint.

This feedback loop is augmented by the use of PID (proportional integral derivative) algorithms.

PID control

In proportional control, actual temperature will approach setpoint, but because of various factors, may not reach setpoint in all instances. The difference is called “offset.”

If offset were constant, the difference could be programmed into the controller. In the real world, however, offsets change over time and with load conditions, so some means of predicting them must be used.

In integral control, the changing amount of offset is calculated by the control algorithm and is added to the setpoint. Integral control is often called “reset” control because of this characteristic.

Derivative control looks at the slope of the curve of the temperature change. If the slope is steep, the algorithm moves the valve faster (or by a greater amount) to meet the new conditions.

Algorithms that use PID control can be very accurate. In many instances the algorithm can be made to “learn” its own coefficients for the three variables.

Autotune PID controllers have this ability; however, they are generally only available for single-point temperature control. Simple evaporator temperature or discharge gas bypass applications may lend themselves to this type of control, but complex functions such as superheat control are usually not available in off-the-shelf components.

Since expansion valves, whether mechanical or electronic, are essentially superheat control devices, controllers for them must measure superheat. This can be accomplished by the two-temperature method or pressure-temperature method.

In either case, extensive testing is required to prove proper operation. (Due to the expense of this testing, superheat algorithms are held secret by most control manufacturers.)

Discharge air or water control

A secondary routine in EEV algorithms may control the temperature of the discharge air or water directly. In this design, as long as superheat remains above some minimum value, temperature of the medium being cooled is the control setpoint.

If superheat falls, the controller will resume superheat control and attempt to raise superheat to the set value. Once superheat is re-established, discharge temperature control is resumed. This type of algorithm may be suitable for some process applications, but has been found to be less desirable in supermarket display cases.

In a refrigerated display case with direct air temperature control, the efficiency of the EEV allows less of the evaporator to be used, but at a higher TD (temperature difference). Higher TDs on the evaporator may lead to an increase in frost and require longer or more frequent defrost periods. In general, EEVs using superheat control algorithms are less likely to build frost.

In systems with coils specifically designed for EEV control, or with provisions to float suction pressures, EEVs may increase control precision while saving energy.

As more systems are designed to take advantage of the unique operating characteristics of EEVs, we will see great improvements in system efficiency, serviceability, and economy.