COMPONENTS OF POWER-SUPPLY SYSTEMS

All the power-supply systems discussed in this chapter are comprised of individual components, or building blocks

The degree of power continuity and overall system re­liability and cost achieved m any system depends on how the basic building blocks are combined The following sections describe the building-block components of the system. Some necessary aspects of properly specifying the components as part of the overall system are also covered.

9- 4.1 Static Inverters

Static inverters become an essential part of a high — reliability power system if short transfer time and an a-c output are required. These inverters are usually used in conjunction with batteries and a static rectifier to provide, on failure of the plant auxiliary-power source, instanta­neous transfer of power from the plant auxiliary-power source to the battery system.

Static inverters consist of three basic parts a low-power oscillator, a power-switching section, and (usually) an output ferroresonant transformer. The low-power oscillator determines the operating frequency of the inverter and can be independent of or synchronized with the plant auxiliary — power system. Once the inverter has been running for several hours, an output-frequency stability within +0.25% of the desired frequency is obtained The major factors tending to change the frequency are long-term drifts in components and variations in the ambient temperature

The power switching section is probably the most critical portion of the overall inverter and usually consists of bridge-connected silicon controlled rectifiers (SCR) The four main SCR’s are alternately switched in pairs, which converts the input d-c source into a square-wave alternating voltage that is applied across the primary of the output transformer The peak amplitude of the alternating voltage is essentially equal to the direct input voltage At the end of each half cycle, the two conducting SCR’s are shut off by momentarily providing a reverse voltage bias This process, called commutation, is an extremely critical function in the proper operation of the inverter If during any half cycle the commutation should fail, the system would be left with more than two SCR’s in the conducting state This would result in an effective short circuit across the battery and would shut down the unit Ensuring proper commutation even under the most adverse conditions is essential in designing reliable inverters In addition to proper commuta tion, the associated circuit must limit the SCR rate of rise of current (di/dt) or voltage (dv/dt) and the peak forward and reverse voltages that appear across the SCR’s

In most applications the desired output is a regulated low-distortion sine wave. Usually both regulation and filtering are provided by the ferroresonant output trans­former, a passive magnetic system similar to the commonly used constant-voltage transformer. An important feature of the ferroresonant transformer is that, as the load current is increased, the output voltage remains essentially constant up to a point in excess of rated load. Above this point the characteristic becomes a very nearly constant current mode As a consequence of this, the inverter can be operated continuously into any overload, up to and including a short circuit, without affecting the square-wave switching portion of the inverter A sine-wave inverter of this type can therefore satisfactorily handle load transients that might otherwise cause misoperation or lack of commutation in the square-wave section

The normal regulation that can be expected is ±3% for all conditions of input voltage and for loads between zero and rated maximum at unity power factor Loads at other than unity power factor have an additional effect on the output Generally, inductive loads reduce the output voltage whereas capacitive loads increase the output volt­age. Loads with a power factor below 0 8 lagging increase the harmonic distortion in the output For these reasons it is preferable to operate with a load having a power factor as near unity as practical, unity power factor also corresponds to minimum d-c input dram. Where the load has an inherent low power factor, the designer must provide suitable correction either at the load or the inverter

In applications where other critical loads are also fed from the battery system, it is desirable to use a filter on the input to the inverter With a sine-wave output from the inverter, the input current resembles a half-wave rectified sine wave Superimposed on this direct current is a large a-c component that may modulate the battery voltage suf­ficiently to cause an undesirable hum in the input of other equipment fed from the battery. The filter eliminates this problem. A second and probably more important function of an input filter is the elimination of spikes that are generated across the battery by other equipment, such as d-c motors and solenoids. Such equipment, commonly used in nuclear power plants, is notorious in generating large voltage transients during operation. Inverter input protec­tion should be provided for short-term transients, in the order of 100/isec up to 4000 volts, when fed from large-station battery systems

Table 9.1—Input and Output Ratings of Typical Static Inverters

Single-phase output

Three-phase output

Output voltage,

120

208Y/120

volts (a-c)

230/399

Frequency, Hz

50 or 60

50 or 60

Output, kVA

2 5 to 28

9 to 150

Input voltage,

48 or 125

125 or 250

volts (d-c)

Output voltage

±1

± 1 to ± 2

and frequency regulation, %

Output harmonic

<5

<5

distortion, %

Inverters are readily available in a variety of standard output and input ratings, the most common of which are listed in Table 9.1. The typical standard ratings in Table 9.1 do not, of course, represent the limits of the manufacturers’ capabilities. Nonstandard inverters of different output and input voltage, kilovolt-ampere rating, and output quality are available for a premium price on request. The following is an outline of the major areas of importance which should receive attention when specifying a static inverter

The range of input voltage over which the required output must be maintained for a stated time during normal and emergency operation is most important In most cases the inverter supplier does not have control over the input source, which is usually the nuclear-power-plant station battery. It is not sufficient just to determine the normal long-term variations of input voltage. Transient input voltages are also important A common design error is to focus on the large-magnitude short-time transients and neglect the higher energy transients of low frequency (0 5 to several Hz). Input voltage transients due to starting motors may not show up on either a long-time source voltage recorder or an oscilloscope set up to detect switching transients Comprehensive knowledge of the characteristics of the input source is prerequisite to proper inverter application and protection

The precise definition of source impedance is not always necessary. However, it is relevant to note the length and size of conductors between the inverter and the d-c source and between the inverter and any switching or protective devices in the incoming lines Because of input-current pulsing dunng the SCR switching, an input filter may be desirable to remove unwanted modulation of the d-c source. This adds to the inverter cost, and, the actual need for it should be determined before specifying an input filter

The load characteristic should be carefully defined Of specific importance are the load power factor, the variation of load, and the maximum load to be switched at one time. Static inverters have definite limits to their momentary overload capacity, and the limits cannot be exceeded. This characteristic is different from the characteristic of a rotating inverter that has inertia and becomes very im­portant where motor loads or other high inrush loads are present. The possibility of load short circuits should also be considered. It may be concluded that a current-limited output is desirable, and, if so, this should be specified

The efficiency of static inverters may be defined in two ways For the amount of heat to be removed due to losses in the inverter, efficiency may be expressed as the ratio (in percent) of output power losses to rated output power losses However, if the purpose is to determine the source current, efficiency is expressed as the ratio of rated output a-c power to rated input d-c power (in percent). It is important to indicate which definition is to be submitted by the manufacturer in his bid proposal This is particularly important for load power factors, which differ significantly from unity.

The output of the static inverter may be synchronized with another source or with a frequency standard. It is important to indicate the impedance, potential variation, and transient noise capability of the synchronizing signal to be used.

In most cases an extremely low harmonic output distortion level is not necessary. If the a-c output is to be rectified and filtered, a square-wave output would even be desirable. Should a square-wave inverter output be used in conjunction with external transformer loads, the trans­formers must be capable of handling the additional 11% swing in flux without overheating.* In general, a wider harmonic distortion tolerance in the specification results in reduced size, cost, and weight of the inverter.