Most power quality disturbances that affect digitally dependant industries are short in duration, so the best option for providing seamless power may be energy storage technologies

Production lines in high-tech manufacturing plants operate through digital switches. Transportation companies rely on digital devices for scheduling and ticketing. Telecommunications facilities require digital equipment to provide customers with the best service. Even financial institutions use binary code to carry out numerous transactions. And despite the technologically advanced nature of all of these industries, each one can be brought to a grinding halt by an event that lasts less than a second.

All of these industries are vulnerable to small power quality disturbances because they rely on digital controls. Even a 30% sag that lasts just a few cycles can close down an entire plant. Power outages are so often associated with massive events like the recent blackout in the Northeast that it's easy to overlook the damage and disruption that short-duration disturbances can cause to facilities that rely on 24/7 power. To make matters worse, recovery from outages can take several hours, making power quality events a costly disruption for any facility. Estimates for costs associated with outages throughout the United States range from $35 billion to $150 billion per year.

Not long ago, power quality was a technical issue that interested only a few highly specialized experts. The rapid rise of digital technology, though, has made it a widespread concern. And as engineers and facilities managers do what they can to increase the “nines” at their plants, the manufacturers of backup power technologies are doing what they can to help out.

Because the majority of power quality events are so short, the most likely choice for combatting the problem is battery backup. Energy storage technologies have made great strides in recent years and now offer a variety of choices for ride-through power, proving that although simple in concept, batteries can provide a reliable solution to a complex problem.

Battery technology. A wide array of batteries is available for energy storage. Dr. Imre Gyuk, program manager for the Energy Storage Program at the U.S. Department of Energy (DOE) in Washington, D.C., breaks down the various types of batteries currently available for purchase into three categories — classical, flow, and advanced — and provides a practical look at how they're applied in real life settings.

Classical batteries comprise the styles and models that are most readily associated with the term “battery.”

Lead-acid (LA) batteries. Widely used in automobiles and trucks, they're inexpensive and their operating characteristics are fairly well known. The largest battery system using LA batteries is located in Puerto Rico and provides spinning reserve as well as voltage and frequency control for the island's power grid. At 20MW and 14MWh, the system delivers both power and energy. Although the tropical location makes extra demands on the batteries, officials have decided to double the capacity of the system in the near future.

A transportable LA system with carefully integrated power electronics has been developed with partial funding from the DOE. The system allows high-tech customers to avoid the disastrous effects of micro outages and voltage sags with a payback of one to two years. It has been used successfully in polymer extrusion factories and semiconductor plants with loads ranging from 2MW to 16MW.

Valve-regulated lead-acid (VRLA) batteries. VRLAs require less maintenance than standard LA batteries. They're sealed, except for a small vent that controls internal pressure, and they require no topping off with water. On the Alaskan island of Metlakatla, engineers installed a 1MW/1.4MWh VRLA system to shield a town from the frequent momentary brownouts that occurred whenever a local sawmill processed a tree. Gel batteries. These batteries are becoming popular in Europe because they're more robust and can take more heat and charge abuse. However, they're more expensive. Nickel cadmium (NiCd) batteries. Despite their rarity in large stationary applications, a 40MW system is being built for voltage support on a long power line to Fairbanks, Alaska. Their resistance to cold may have been among the deciding factors in selecting this type of battery.

Flow batteries present an interesting feature: they can decouple power and energy. A central cell stack provides power, but total energy is furnished by a reservoir of rechargeable electrolyte, which can be situated anywhere that's convenient and for which there are no size limits.

Zinc-bromine batteries. These are available off the shelf and have been deployed at a number of sites. Again, integrated power electronics are essential to successful applications. Vanadium redox batteries (VRBs). Developed in Australia and Japan, this technology has been deployed in units as large as 500kW/10 hrs for load management in Japan. One of the more interesting applications features electrolyte storage in plastic bags, which can be stuffed into available crawl spaces and other residual areas. The technology also can provide voltage control. Sodium bromide batteries. These batteries have received considerable interest recently. A 15MW/8 hr facility is under construction in the United Kingdom. A similar system is planned for the state of Mississippi in collaboration with the Tennessee Valley Authority (TVA).

Advanced batteries offer a vastly decreased footprint and excellent maintenance characteristics when compared to the two other categories of batteries. However, they tend to be expensive for large-scale applications. Lithium ion, lithium polymer, and nickel metal hydride batteries have been developed mainly for automotive use. Sodium sulfur batteries are a special case among advanced technologies.

Developed in Japan, this battery operates at high temperatures. Extensive tests have demonstrated safe containment under extreme conditions. Some 38 systems totaling about 2MW and 124MWh have been installed in Japan. The largest of these installations, located at a substation in the foothills of Mt. Fuji, provides 6MW for 8 hrs. It also can supply active and reactive power to mitigate voltage sags and frequency fluctuations. Operation of a 500kW unit should begin in the United States in the near future. The unit will be used for load leveling or as an uninterruptible power supply (UPS).

Flywheels and more. Whereas classical LA batteries or novel flow batteries rely on chemistry, flywheels, supercapacitors, and superconducting magnetic energy storage (SMES) systems rely on physics. Therefore, they're more robust and less subject to the ravages of entropy. Flywheels store kinetic energy, capacitors store electrostatic energy, and SMES systems store magnetic-field energy. A GE EDG presentation titled, “Advances in Energy Storage for Critical Power Systems” provides insight into these evolving energy storage technologies.

Flywheels. Although an integral part of engine designs for hundreds of years, flywheels are increasingly attracting interest in the power quality arena. In critical power systems, flywheel energy storage systems (FESS) are used to bridge short-term (5 sec to 30 sec) power quality events or to support the load until a fast-start backup generator is activated. Fig. 3 (above) shows a simplified representation of a FESS. Fig. 4 shows a cross section view of one type of flywheel, along with an elevation of its free-standing enclosure with components noted.

According to the presentation noted above, increasing the radius or speed of a flywheel can increase the stored energy. However, this will also increase the internal stress on the material and/or require more sophisticated bearings. As a result, there are two types of flywheels currently on the market: high-speed flywheels (25,000 rpm to 80,000 rpm) made of composite material and low-speed flywheels (2,000 rpm to 10,000 rpm) made of steel.

In a high-speed FESS, a power electronic rectifier/inverter, which is necessary to correct the large volume swing at the generator terminals over the speed range, is connected at the output of the generator. The energy is injected at this point into the DC link circuit of the system. The high-speed FESS has the advantage of compact size for a given amount of power delivered.

On the other hand, low-speed systems are considerably larger than their high-speed counterparts for a given amount of stored energy, their flywheels are made of steel instead of composite materials, and they use conventional bearings with a magnetic support, which is used to support the large rotor weight. Only a partial vacuum is necessary to maintain low frictional losses. These losses, combined with the excitation losses, result in a float charge that's roughly double that of typical high-speed storage systems.

Like the high-speed FESS, most low-speed FESSs use either a single-stage insulated gate bipolar transistor (IGBT) or a rectifier/inverter for power transfer. Some operate with an induction or mechanical coupling instead of this DC interface. The former coupling type can be used to correct for changing rotational frequency. The latter coupling type, however, must compensate for this through power electronics that act as a 3-phase exciter winding in the generator.

The greatest advantage that flywheels offer over batteries is their ability to predict with pinpoint accuracy the amount of energy remaining.

High-temperature superconducting flywheels are currently under development with funding from the DOE. Such systems would offer inherent stability, minimal power loss, and simplicity of operation, as well as increased energy storage capacity.

Supercapacitors. Also known as ultracapacitors, these unique devices — basically electrochemical double-layer capacitors — consist of two electrodes, a separator, and an electrolyte. The electrodes can be made from a highly conductive alloy containing metal oxides, carbons, and conducting polymers. The separator is a membrane that allows mobility to charged ions but restricts electronic conductance. The electrolyte may be solid state, organic, or aqueous, depending on the type of supercapacitor and application.

Like a battery, the operating voltage of a cell is about 1.3V to 2.5V. However, combined parallel strings of supercapacitors can achieve hundreds of volts. Since these values are typical for flywheel systems, many compare supercapacitors with uncoupled flywheel systems. A comparison of the advantages and disadvantages of each is shown in the table).

Supercapacitors show great promise for industrial applications because of the following attributes:

  • Technical maturity

  • Chemical inertness

  • Very limited environmental impact (no use of heavy metals)

  • Expected performance vs. reasonable cost

Although these attributes are impressive, a large-scale supercapacitor system has yet to be fielded.

Superconducting magnetic energy storage systems (SMESs). Instead of storing kinetic or chemical energy, SMES devices store energy in the magnetic field generated by a loop of endless current ( Fig. 5 below) A zero-resistance coil passes DC current and stores in its magnetic field an amount of energy described by the equation E=0.5×L×I 2. Superconductivity allows the coil to be wound very compactly, resulting in an achievable high specific energy density. Obviously, the superconductivity also minimizes resistive losses of the passing current. Since the instantaneous energy content depends on the square of the current, the remaining energy can also be measured very accurately.

The advantage of the SMES lies in its ability to store large amounts of power and to charge and discharge orders of magnitude more times than even the best batteries. Also, it has high efficiency for short-duration storage, making it well suited for pulse discharge applications.

Unfortunately, the effect of SMES technology on light industrial or UPS applications isn't very clear. SMES-based systems tend to be more expensive and require larger space than other comparable energy storage systems. Also, they are less suited to continuous standby applications. Nevertheless, the ability to operate at low- and medium-voltages and discharge up to 3MW from a single unit makes this technology appealing for very critical loads.

Several 1MW units are used for power quality control throughout the world. One recent development was the deployment of a string of distributed SMES units in northern Wisconsin to enhance stability of a transmission loop. The line had been subject to large sudden load changes due to the operation of paper mills and had the potential for uncontrolled fluctuations and voltage collapse. Besides stabilizing the grid, the six SMES units also provide increased power quality to customers served by connected feeders. Other units have recently been installed in the Houston area, and the TVA is installing a unit for fast power control.

Fuel cells. Believe it or not, the first fuel cell was developed in 1839. Primitive in comparison to today's design, Grove's invention was the first in a long line of stored energy devices. However, fuel cell technology wasn't put into use until the '60s, when NASA installed this technology to generate power on the Gemini and Apollo spacecraft.

The above-mentioned paper also provides insight into the technological advances in this technology. Fuel cells, an alternative to reciprocating engines, are electrochemical devices that directly convert hydrogen or hydrogen-rich fuels into electricity without combustion (Fig. 6). The process is very efficient, with the cells theoretically converting as much as 80% of the chemical energy in the fuel into electricity.

These devices are similar to a battery in structure, with two porous electrodes separated by an electrolyte. They use a chemical reaction between a hydrogen-based fuel and an oxidant, usually oxygen, to produce electricity. The hydrogen-based fuel is fed into the anode of the fuel cell. Oxygen then enters the fuel cell through the cathode and, with the aid of a catalyst, the hydrogen atom splits into a proton (H+) and an electron. The proton then passes through the electrolyte to the cathode, and the electrons travel through an external circuit connected as a load and create a DC current. At the cathode, the protons combine with hydrogen and oxygen to produce water and heat.

The general design of most fuel cells is similar, except for the electrolyte. There are five types of fuel cells, each defined by its electrolyte:

  • Alkaline

  • Solid polymer (proton exchange membrane, or PEM)

  • Phosphoric acid

  • Molten carbonate

  • Solid oxide (SOFC)

Alkaline and PEM fuel cells operate at lower temperatures (50°C to 260°C) and are designed mainly for use in transportation applications. The other three types operate at higher temperatures (as high as 1,000°C for SOFC) and are being developed for use in co-generation and large central power plants. Fuel cells generate very low levels of NOx and CO emissions because the power conversion is an electrochemical process.

The stack, or the part of a fuel cell that contains the electrodes and electrolytic material, is the major contributor to the cost of the total system. Although costly, stack replacement is necessary when the fuel cell's efficiency degrades and as stack operating hours add up.

What the future holds. There are a number of discernible trends in energy storage technology. According to Dr. Gyuk, storage units, particularly those used in applications requiring appreciable amounts of power, are increasingly factory assembled and tested. Customers want plug-and-play units. They have neither the time nor the expertise to tease a system into optimal performance.

He goes on to say that the trend is toward fully integrated electronics with carefully designed inverters and control systems. Eventually, you can expect the controls to acquire a certain amount of intelligence to respond to weather or market signals. At the same time, such systems will be addressable through the Internet for diagnostics or inventory.

Also, hybrid systems that combine energy storage for seamless response with backup power for reliability are becoming available. Backup usually consists of diesel generators, but microturbines and fuel cells will play a role in the future.

Finally, Dr. Gyuk says that energy storage systems are getting bigger. Reliability for semiconductor factories and other sensitive industries requires multi-megawatt power, and transmission stability applications have even higher needs. Energy management applications may eventually result in storage facilities of greater than 100MW.

Regardless of what method of energy storage you choose, it's becoming clear that some form of backup power will be necessary at most facilities now that digital power needs have become so prevalent. When even a few seconds of downtime can be costly, providing the proper short-duration ride-through capabilities can be a valuable step to protecting critical power processes.




Sidebar: Tale of the “Nines”

The power quality industry expresses the degree of reliability as the number of “nines.” The U.S. national power grid yields about 3 nines, or 99.9% reliability (The Northeast blackout of August 14, 2003 withstanding). But digital industries would like to have no more than a single 1-cycle outage per year, which corresponds to 9 nines of reliability.

This terminology can be somewhat misleading in that the number of nines relates to total electrical supply outages. However, the important figure for facility personnel is total downtime because this number more closely relates to the number of power quality events than to cumulative duration. Equally misleading, at least for digital industries, are the tables that give the cost per hour of a power outage. Recovery time is much the same for 5-cycle outages as for 5-min ones.

A better way to approach this is to look at scatter plots that display the magnitude and duration of voltage fluctuations. Events falling outside the ITIC Curve (see Fig. 1 ) are likely to cause system failures.

EPRI has collected extensive data, and a new study of power quality in Silicon Valley is underway with DOE funding. Fig. 2 (right) summarizes the results of these studies using the framework of IEEE Standard 1159, “IEEE Recommended Practice for Monitoring Electric Power Quality.” The diagram shows the percentage of fatal power quality events in different regions of the chart. The point to note here is that 98% of fatal power quality events are shorter than 15 sec — and most of them are voltage sags.

Energy storage represents virtually the only option for providing seamless continuity of a power supply and maintaining it for the crucial 15 sec. Even the fastest diesel generators can't respond fast enough, and other forms of distributed generation (DG), such as microturbines and fuel cells, are even slower.

Of course, if you have critical loads, you could opt to disengage from the grid and rely entirely on DG. Using an N+1 or N+2 configuration would provide continuous high-quality power, but this option could be considerably more expensive than grid power. It may also conflict with emission standard requirements in many parts of the country. Furthermore, if loads are variable, you will need some energy storage because most DG has bad load-following characteristics.

The ideal solution combines energy storage for at least 15 sec and DG for backup. The price of the equipment, maintenance and fuel costs, and the cost of downtime will determine the exact balance of storage and DG.