This article provides installers of drives with some guidelines on where and when reactors are needed, and on what size reactor to use.
A reactor is essentially an inductor. Physically, it is simply a coil of wire that allows a magnetic field to form around the coil when current flows through it. When energised, it is an electric magnet with the strength of the field proportional to the amperage flowing and the number of turns. A simple loop of wire is an air core inductor. More loops give a higher inductance rating.
Some ferrous material such as iron is sometimes added as a core to the winding. This has the effect of concentrating the lines of magnetic flux, thereby making a more effective inductor.
Going back to basic AC circuit theory, an inductor has the characteristic of storing energy in the magnetic field and is reluctant to a change in current. The main property of a reactor is its inductance and is measured in henrys, millihenrys or microhenrys. In a DC circuit (such as that of the DC bus in an AC drive), an inductor simply limits the rate of change of current in the circuit since current in an inductor wants to continue to flow at the given rate for any instant in time. That is to say, an instantaneous increase or decrease in applied voltage will result in a slow increase or decrease in current.
Conversely, a corresponding voltage will be induced if the rate of current in the inductor changes. If we look at the equation V = Ldi/dt for an inductor where V is voltage, L is inductance and di/dt is the rate of change of current in amp per second, we can see that a positive rise in current will cause a voltage to be induced.
This induced voltage is opposite in polarity to the applied voltage and proportional to both the rate of rise of current and the inductance value. This induced voltage subtracts from the applied voltage, thereby limiting the rate of rise of current. This inductance value is a determining factor of the reactance. The reactance is part of the total impedance for an AC circuit. The equation for the reactance of an inductor is:
XL = 2 π fL
where:
XL = inductive reactance in Ohm.
f = applied frequency of the AC source.
L = inductance value of the reactor.
The reactance and, therefore, the impedance of the reactor is higher with a higher inductance value. Also, a given inductance value will have a higher impedance at higher frequencies. We can therefore say that, in addition to limiting the rate of rise in current, a reactor adds impedance to an AC circuit proportional to both its inductance value and the applied frequency.
Side-effects of adding a reactor
Like most medication, there are side-effects to using a reactor. Though these issues should not prevent the use of a reactor when one is required, the user should be aware of and ready to accommodate these effects. Since a reactor is made of wire (usually copper) wound in a coil, it will have the associated losses due to wire resistance.
Also, if it is an iron core inductor (as in the case of most reactors used in power electronics), it will have some “eddy current” loss in the core due to the changing magnetic field and the iron molecules being realigned magnetically. In general, a reactor will add cost and weight, require space, generate heat and reduce efficiency.
Fig. 1: A standard 3-phase input converter using six SCRs or six diodes and a filter capacitor bank; with a line reactor, and with “DC link choke”.
The addition of a line reactor can sometimes change the characteristics of the line you are connected to. Other components such as power factor correction capacitors and stray cable capacitance can interact with a line reactor, causing a resonance to be set up. AC drives exhibit a relatively good power factor and do not require the use of correction capacitors. In fact, power factor correction capacitors often do more harm than good where AC drives are present. For the most part, power factor correction capacitors should never be used with a drive.
You may find that the addition of a reactor completes the required components for a line resonance where none previously existed, especially where power factor correction capacitors are present. In such cases, either the capacitor or the inductor must be removed.
Furthermore, reactors have the effect of dropping some voltage, reducing the available voltage to the motor and or input of the motor drive. But why use a reactor with all these side effects? If you ask that question, you may hear a whole slew of answers ranging from, “That’s the way we always do it” to “I’d rather be safe than sorry”.
The fact is, there are good reasons to install a reactor under certain conditions. Let’s start with the input side of a drive.
A reactor at the input
As you may already know, most standard “six-pulse” drives are nonlinear loads. They tend to draw current only at the plus and minus peaks of the line. Since the current wave-form is not sinusoidal, the current is said to contain “harmonics”. For a standard 3-phase input converter (used to convert AC to DC) using six SCRs or six diodes and a filter capacitor bank as shown in Fig. 1, the 3-phase input current may contain as much as 85% or more total harmonic distortion. Notice the high peaks. If a line reactor is installed, the peaks of the line current are reduced and somewhat broadened. This makes the current somewhat more sinusoidal, lowering the harmonic level to around 35% when a properly-sized reactor is used. This effect is also beneficial to the DC filter capacitors since the “ripple current” is reduced. The capacitors can be smaller, run cooler and last longer.
Though harmonic mitigation is an important reason to use a line reactor, most drives at the 10 hp rating and above include a “DC link choke” (seen in Fig. 1). The link choke is a reactor put in the DC bus between the rectifier bridge and the capacitor bank. It can provide the necessary harmonic mitigation and, since it is in the DC bus, it can be made smaller and cheaper than the 3-phase input reactor.
Fig. 2: A group of 5 – 10 drives connected through one large 3-phase reactor.
Small drives may need an input reactor
Drives under 10 hp generally do not have a DC link reactor and, in most cases, that’s not a problem since any harmonic current distortion would be small when compared to the total load of the facility. An input reactor is a valid method in reducing harmonics if many small drives are required for a process. In the case of many small drives, it is often more economical and practical to connect a group of five to ten drives through one large 3-phase reactor, as shown in Fig. 2.
If there was ever a mandate to install an input reactor, it may be on a small drive where the transformer feeding it may be 20 times or more of the current or power rating of the drive. In some cases, a large transformer (one with a low source impedance and/or high short-circuit ability) feeding a relatively small drive can result in overheating the drive internal DC capacitor bank.
When a negative temperature coefficient (NTC) pre-charge system is used, a large transformer feeding the drive can result in excessive inrush and clear line fuses or damage the drive. An input line reactor here will help. In this case, the reactor reduces harmonic current but the real reason for its presence is to limit the peak current that will flow at the input and in the capacitor bank.
Fig. 3: A reactor at the input to the drive may be used but a better solution would be to attenuate the voltage spike at the source with a snubber circuit.
A reactor as a line voltage buffer
In some cases, other switchgear such as contactors and disconnects on the line can cause line transients, particularly when inductive loads such as motors are switched off. In such cases, a voltage spike may occur at the input to the drive which could result in a surge of current at the input. If the voltage is high enough, a failure of the semiconductors in the DC converter may also result.
A reactor is sometimes used to “buffer from the line”. While a DC link choke, if present, will protect against a current surge, it cannot protect the converter from a voltage spike since a link choke is located after the converter (see Fig. 1). The semiconductors are exposed to whatever line voltage condition exists. For this reason, a reactor at the input to the drive may be of some help, but a better solution would be to attenuate the voltage spike at the source with a snubber circuit. Fig. 3 shows both methods used to protect the drive input semiconductors. A reactor does not fix grounding issues, nor does it provide isolation. Bear in mind that, while a reactor provides some buffering, it does not provide isolation and cannot take the place of an isolation transformer. An isolation transformer must be used where isolation is needed. While a reactor can provide light buffering from a short-duration (less than 1 ms) transient condition, it will not fix a high line condition or protect against line swells (high line for several line cycles). Nor should it be expected to protect against high energy short-duration events such as lightning strikes.
Increasing load inductance
Applying a reactor at the output of a drive is sometimes necessary. All the side-effects as outlined previously hold true. There are a few instances when it may be necessary to add load impedance by inserting an output reactor. If the motor has a “low leakage inductance”, a reactor can help bring the total load inductance back up to a level that the drive can handle. In the days of the “bipolar transistor” drive, carrier frequencies rarely exceeded
1,5 kHz. This meant that the transistor “on time” was much longer and allowed current to ramp up higher, limited by the load or motor inductance.
The result of a low inductance motor was huge ripple current that sometimes ran into the current limit of the drive, causing poor performance or tripping. For the most part, the higher carrier frequencies and, correspondingly, lower ripple current of today’s Isolated Gate Bipolar Transistor (IGBT) drives have eliminated the need to add inductance to the load (see Fig. 4).
In some rare cases where a strange motor configuration or a motor with six or more poles is used, the motor inductance may be too low and a reactor may be needed. Running multiple motors on one drive may also result in a low inductance load and the requirement of an output reactor.
Reducing the effect of reflected wave
A reactor at the output of a drive is sometimes installed to prevent a reflected wave voltage spike when long motor leads are required. This is not always a good practice. Though the reactor will slope off the voltage rise time, providing some benefit, it is not likely to limit the peak voltage at the motor. In some cases, a resonance can be set up between the cable capacitance and reactor which causes even higher voltages to be seen at the motor. In general, a motor terminator is a better solution. If a reactor is installed at the output, it is most likely part of a specially-designed “reflected wave reduction” device which also has damping resistors in parallel.
If a reactor is used at the output, it should be located as close to the drive-end as is possible. Fig. 5 shows the motor voltage before and after the installation of a reactor. The DC bus voltage is shown for reference. Notice that the rise times are different, the peak voltage is about twice the DC bus voltage regardless of the use of a reactor.
Since a current-regulated drive requires “voltage margin” to regulate current, the output voltage is already limited by about 5%. Adding a reactor at the output will drop the voltage even further. A reactor at the output of this type of drive may not be a problem as long as the application can run without full motor voltage near full speed (typically 55 to 60 Hz). In some cases, a specially-wound motor may be used to compensate. For example, a 460 V, 150 A motor may be rewound as a 400 V, 175 A motor.
Fig. 4: The need to add inductance to the load has been eliminated.
Sizing a reactor
The first rule is to ensure that you have a high enough amp rating. In terms of the impedance value, you will usually find that 3 – 5% is the norm with most falling closer to 3%. A 3% reactor is enough to provide line buffering and a 5% reactor would be a better choice for harmonic mitigation if no link choke is present. Output reactors, when used, are generally around 3%. This percentage rating is relative to the load or drive where the reactor impedance is a percentage of the drive impedance at full load. A 3% reactor will therefore drop 3% of the applied voltage at full-rated current. To calculate the actual inductance value, we would use the formula:
L =XL/(2 π fL)
where:
L = inductance in henrys.
XL = inductive reactance or impedance in Ohm.
f = frequency.
Fig. 5: Motor voltage with and without a reactor.
The frequency will generally be the line frequency for both input and output reactors. Your drive distributor should be able to help you size a reactor for use with a drive. The following example may be helpful if you wish to calculate the value yourself:
If a 3% reactor was required for a 100 A, 480 V drive, a 100 A or larger current rating would be required.
The drive impedance would be Z = V/I or 480/100 = 4,8 Ω. 3% x 4,8 Ω =
0,114 Ω. Inserting this 0,114 Ω impedance in the equation for inductance, we get a value of about 300 microhenrys.
Summary
A reactor is no silver bullet but it can prevent certain problems when applied properly. For the most part, a reactor at the input or output is not required automatically. Reactors can be helpful in providing some line buffering or adding impedance, especially for drives with no DC link choke. They may be needed for small drives to prevent inrush or to provide reduction in current harmonics when many small drives are located at one installation. At the output, they should only be used to correct low motor inductance and not as a motor protection device.
Use a reactor to add line impedance; provide light buffering against low-magnitude line spikes; reduce harmonics (when no link choke is present); compensate for a low-inductance motor, and only as part of a filter for reflected wave reduction.
Contact Michelle Junius, Rockwell Automation, Tel 011 654-9700, mjunius@ra.rockwell.com
The post Line reactors and AC drives appeared first on EE Publishers.