The number of electrical devices damaged or destroyed by surge voltages is increasing year by year. This can prove expensive in terms of repairs and downtimes. In an industrial environment, the hazards are not only restricted to systems and devices. Depending on their level, the effect of surges reaching electronic equipment ranges from operational malfunction to complete failure. At some level, the surge may only cause temporary upset with no permanent hardware effect, but at higher levels, it may cause failures of critical components.
Interference voltages: Switching operations triggered mechanically or electronically generate pulse-like and high-frequency interference voltages. These voltages spread in an unimpeded manner across the cable network. All the devices within this cable network are affected. Data errors, uncontrolled functions, and system crashes can result with electronic and data processing devices at particular risk.
Lightning discharge: Lightning strikes are originated by the electric charges accumulated in the clouds: the strike is the actual electric breakdown between the cloud and ground. The discharge drives a surge current of various kA, this current being driven similarly to that from a current source. The resulting waveform is typically an impulse, with a front as short as a microsecond (time to peak: 0.1 µs < T < 20 µs), lasting a few to various microseconds (tail duration < 300 µs), involving frequencies of hundreds of kHz (10 kHz to 1 MHz). Lightning surges are a major issue, and may be causing faults or failures throughout the power system. In fact, the most stressful transients in a power system are due to lightning strikes to the transmission lines, and associated flashover. Particularly, lightning is considered the greatest single cause of line outages. It is above all lightning strikes i.e lightning electromagnetic pulse that has the greatest potential for damage among all the causes of occurrence. They cause transient overvoltages that can extend across great distances and are often associated with high-amplitude surge currents. Even the indirect effects of a lightning strike can lead to a surge voltage of several kilovolts and result in a surge current of tens of thousands of amperes. In spite of the very brief duration, such an event can lead to total failure or even the destruction of the entire system.
Switching operations: Switching operations i.e. switching electromagnetic pulse can generate induced surge voltages that spread to supply lines. In the case of large switch-on currents or short circuits, very high currents can flow within a few milliseconds. These short-term current changes can lead to transient overvoltages.
Electrostatic discharge: Electrostatic discharge occurs if exposed conductive parts with different electrostatic potential approach each other and result in a charge exchange. A sudden charge exchange leads to a brief surge voltage. This presents a hazard, especially, for sensitive electronic components.
Regardless of what causes a surge voltage the consequences are the same- device destruction, system downtimes and total failure of controllers. Device failure or defects caused by surge voltages are more frequent than expected. Many a time, the consequences of a failure are generally much more serious such as downtimes or data loss. The failure of a device or a machine that is used in a professional environment often leads to costs that are many times higher than repairing the defective device.
Effective surge protection starts with assessing the potential risk and identifying all the devices within the item to be protected. The resulting protection concept takes into account all the interfaces of the power supply unit as well as those for data and telecommunications. This is the only way to comprehensively and effectively protect all the terminal devices, for example, within a data network, production plant or building installation. Combining high-quality protective devices with innovative arrester technology, surge protective devices ensure a high degree of system availability and safety in all areas of electrical engineering.
Lightning can strike more than once, and surges don’t always come from outside. Surge events, short-term transients in voltage threatening critical downstream equipment happen for many reasons. The most common source, though, is internal devices powering on and off: motors, transformers, photocopiers, fluorescent lighting ballasts, light dimmers, variable frequency drives and more. They can also be generated externally by events like lightning, grid switching or electrical equipment in adjacent buildings. While seemingly surge events can wreak serious havoc on unprotected and inadequately protected facilities. They can disrupt or destroy sensitive microprocessor-based devices like computers, programmable logic controls, etc., resulting in premature aging of equipment, process interruptions and catastrophic failures. This change requires surge protection to be installed on all emergency electrical equipment to improve the reliability of emergency power systems. In power converters, fast-fronted transients may force a change of state of power semiconductor switches. Particularly, Silicon Controlled Rectifiers or thyristors can be spuriously fired with high dv/dt surges. Similarly, it is also possible to switch on an Insulated Gate Bipolar Transistor with high dv/dt surges.
Uninterruptible Power Supplies (UPS) provide power conditioning and backup power for critical and sensitive equipment, the typical example being Information Technology (IT) equipment in data centers. Various UPS topologies provide different level of power quality to the critical load. However, all UPS topologies feature unprotected paths that may expose the load to transient overvoltage i.e. fast disturbance, either in the form of impulse or ringwave, resulting in significant overvoltage, from the incoming utility. In this context, double-conversion topology provides the best protection, and it is often the preferred choice for larger UPS installations. Various double-conversion UPS can operate in a high-efficiency mode, typically referred to as ‘ECO’ mode, where the load is normally fed by the bypass path. However, market adoption of ‘ECO’ like operating modes has been poor, following the concern that bypass operation may expose the load to the utility power disturbances, and particularly, transient overvoltage. IEC standard 62040-3 specifies three main types of UPS i.e. passive stand-by, line-interactive and double-conversion.
Passive Stand-By: It is the simplest topology for UPS systems, widely used in low-power applications, upto 2 kVA e.g. protection of a single workstation. The UPS normally operates offline, with the load being fed directly from the utility line. In normal conditions, the inverter is off while the charger is re-charging the batteries. In case of a power failure, the UPS switches to inverter dragging power from the battery stored-energy mode. This topology provides no line conditioning when operating on mains, although it may include an input filter, filtering utility noise and surge circuitry, providing limited surge and spike protection.
Line-Interactive: This topology is popular for applications upto 5 kVA. Similarly, to the stand-by UPS, a line-interactive unit can power the load directly from the utility. However, this topology features a parallel converter that may provide power conditioning by interacting with the utility, additionally; the converter re-charges the battery in normal mode. In case of a power failure, the utility is isolated via the static switch and the converter operates as an inverter feeding the load from the batteries, stored energy mode. This topology typically includes some sort of surge protection e.g. transient clamping components.
Double-Conversion: This topology is the preferred choice for larger installations for UPSs larger than 5 kVA. In normal mode, figure 1, the load is fed via the rectifier-inverter path. The inverter produces a regulated AC output, with voltage and frequency controlled at all times independently from the quality of the AC input. In case of a power outage, the inverter drags power from the batteries. This UPS topology provides the best protection for the critical load. Particularly, transformer-based units prevent the propagation of any disturbance from the input to the output bus.
In normal operation, the load is fed through the rectifier or inverter path. The rectifier converts the AC input into a regulated DC voltage, providing a DC feed to the inverter while recharging the batteries. Then, the inverter converts the DC voltage into a fully regulated AC output. In case of a power outage, the inverter is fed by the batteries (stored energy operation). Additionally, double-conversion UPS often feature a bypass path, with a static switch connecting AC input and output, allowing bypass operation, load fed directly by the AC input utility. However, this mode is mainly used for emergency operation or during maintenance.
Following the AC-DC-AC conversion, output voltage and frequency are controlled at all times, and they are independent from the quality of the AC input supply. Therefore, this UPS topology provides top protection for the critical load. Particularly, transformer-based units also provide input to output galvanic isolation during double-conversion operation, preventing the propagation of any disturbance from the AC input to the output AC bus.
While providing optimum protection, double-conversion UPS offers limited efficiency. In fact, best-in-class efficiency for transformer-based double-conversion UPS is around 94 per cent. In order to overcome this limitation, some double-conversion UPS feature a high-efficiency ECO mode, where the load is fed directly by the input utility via the bypass path, as long as utility remains within given tolerances. The inverter is maintained in a stand-by state, ready to take on the load whenever the bypass utility experiences a disturbance. This operating mode greatly reduces losses, with resulting efficiency exceeding 98 per cent.
Advanced ECO Mode operation is particularly critical, as the load is normally fed directly by the input utility. This may prompt concerns than bypass operation may expose the load to the utility power disturbances, and particularly, lightning surges. However, basic surge protection and transient filtering may be implemented on the UPS. Here, the rectifier section typically includes an input filter. While this filter is mainly aimed to improve the rectifier input characteristics in terms of harmonics, it also constitutes a filter that may effectively attenuate transient overvoltage. In this operating mode, load-sharing between parallel units is not actively controlled. To limit the load unbalance due to cable length differences, an inductor is added to the bypass line, resulting in series with the load. Additionally, the inverter output filter is energised by back-feed from the bypass utility via the UPS output, resulting in parallel with the load. Therefore, Advanced ECO Mode operation combines a bypass inductor placed in series with the load, with the inverter filter capacitors placed in parallel with the load. Now, the combination of a series inductor and parallel capacitor provides some protection against surges, ideally complementing external surge protection devices. As a matter of fact, the bypass inductor in conjunction with the inverter output capacitors constitutes an L-C filter, for which the resonant frequency is given by: fr =1/2π√LC. It is clear that the selection of inductor and capacitors is driven by their main function and not by the desired filtering frequency. However, in most application the resonant frequency would fall in the 1-2 kHz range.
All UPS topologies may expose the load to transient overvoltage from the incoming utility. In fact, stand-by and line-interactive UPS feed the load directly with the input utility, and this is also the case for double-conversion systems during bypass operation. For this reason, the IEEE ‘Emerald Book’ recommends Surge Protection Devices (SPD) to be installed at the UPS input.
Typically, surge protection includes some clamping devices, such as metal oxide varistors (MOV). However, these devices tend to exhibit significant leakage current at low voltages. This characteristic drives the device selection depending on the continuous operating voltage to which they are exposed, resulting in a clamping voltage that may exceed 200 per cent of the nominal utility voltage. This protection may result insufficient for IT equipment.
Dr Gopalkrishna Dhruvaraj Kamalapur
Professor, Department of Electrical and Electronics Engineering
Shri Dharmasthala Manjunatheshwar College of Engineering and Technology, Dharwad
