Selecting the Right UPS for the Job

When most people think about poor power quality, the first things that come to mind are blackouts or brownouts. In fact, these are some of the most uncommon power quality issues facing the industry today. The most typical power quality problems are undervoltage (sags), spikes and surges, overvoltage, noise, and lastly, blackouts. While outages are the most visible of power problems, equipment damage, data loss, and spurious errors are usually caused by these other more common power quality problems.

The standard for “quality power” is largely defined by CBEMA (Computer and Business Equipment Manufacturers Association). The CBEMA curve, as shown in Fig. 1, is a “power quality envelope,” illustrating the acceptable undervoltage and overvoltage conditions that most equipment can sustain for a period of time. For durations longer than one second, CBEMA-compliant equipment can tolerate up to 106% overvoltage and 87% undervoltage conditions. Between 30% to 50% of equipment in today's office and manufacturing environments requires “quality power” in the CBEMA range or better to avoid malfunction, equipment damage, data loss or other critical power-related failures.

The cost of critical power protection can pay for itself in just about all cases with the average small- to medium-sized business losing about $1,500 per year in productivity due to power-related equipment failures. In the case of many continuous-process plants, hospitals and data centers, the cost of a power failure can be exponentially higher. The growing emphasis on plant floor to boardroom data flow for enterprise resource planning (ERP) systems also relies on dependable power. In fact, the wired enterprise's need for continuous power to support information technology now demands 24×7 hours operation at a level known as five-nines reliability (0.99999), which means in any single year the sub-power quality is below acceptable levels for less than six minutes.

For the plant engineer or manager, 0.99999 power reliability extends far beyond the capabilities of local electric utilities. Ongoing deregulation of utilities as well as a growing demand for power makes this situation potentially more troublesome. Uninterruptible power supplies and generator systems will require greater attention from plant engineers than in years past as the digital age unfolds.

There are several considerations when deciding on what type of UPS to use. The major options include using a large centralized three-phase UPS or a smaller-application specific single-phase UPS. To decide what type of UPS is required for an application, consider the following:

How important are the loads (non-essential vs critical loads)?
Which UPS topology best supports the applications (online, line interactive, off-line)?
What are the power requirements (is three-phase power in use)?
How difficult is UPS installation?
When is maintenance required and how extensive is that maintenance?
What is the life-cycle cost of the system?

Non-Essential Vs Critical Load

The consequence of equipment failure is the single most important factor in deciding on the type of UPS. For small office PC loads where single station or small network data protection is required, a small single-phase UPS is often an adequate solution. Most single-phase UPSs use off-line or line-interactive topologies, resulting in a small (but acceptable for standard PCs and servers) interruption when transferring to and from battery power. Because PCs, workstations and peripherals are often located in a decentralized manner throughout offices, a dedicated off-line UPS often priced under $100 provides an ideal solution.

If the equipment to be protected is critical, an online UPS is the best choice. This topology transfers to and from battery and internal bypass without any power interruption to the load, providing a truly seamless transfer to battery power.

For loads above 10kVA, the most practical solution is a three-phase UPS, which in most cases is designed with true-online topology. Three-phase online UPSs offer the advantage of providing centralized protection using a single UPS. This simplifies maintenance and battery replacement while supplying high quality uninterruptible power to critical loads. Large server farms, data centers, 7×24 offices and critical process plants with higher loads are best served with single or parallel three-phase UPSs, often with a redundant UPS backing up the primary UPS.

UPS Topology Choices

The power protection technology supporting desktop computers and workstations differs significantly from the advanced power management systems that protect mission-critical equipment. This creates a lot of confusion in trying to understand the three basic types of UPSs:

Off-line UPS (system or data loss is an inconvenience).
Line-interactive UPS (system or data loss is a serious problem).
Online UPS (system or data loss is unacceptable).

The following discussion provides an overview of the various types of UPS technologies.

Off-Line UPS

UPS with offline topology is shown in Fig. 2. The inverter is connected in parallel and acts simply to backup utility power. The first component in an off-line UPS is input surge protection to protect the load from high-voltage surges/spikes. The second element is the battery to supply the inverter with power, and the third element is the inverter. The inverter takes the dc battery voltage and creates the ac voltage required to power your equipment.

The fourth component is the battery charger. Under normal conditions, the inverter is sitting idle until the input voltage goes above or below a usable level. At that point, the inverter will turn on and supply the load with ac power.

Line-Interactive UPS

With the line-interactive topology, as shown in Fig.3, the inverter is connected in parallel and acts to backup utility power. It also charges the battery. Through its reversible operation, it interacts with utility power to stabilize the voltage.

A line-interactive UPS is similar to an off-line UPS because when sensing an undervoltage or overvoltage situation, it also requires a transfer time for the inverter to turn on and to supply power to the load. The main difference between an off-line and a line-interactive UPS is that a line-interactive UPS in the stand-by mode has active voltage regulation. The primary advantage of a line-interactive over an off-line UPS is that it doesn't use its batteries nearly as often, which extends battery life.

Online UPS

As depicted in Fig. 4, an online UPS offers double-conversion topology: The inverter is connected in series between the ac input and the load. Power for the load flows continuously through the inverter. An online UPS has some of the same components as an off-line UPS with a few differences — most notably, a rectifier. An online UPS has input surge protection, batteries, inverter components and a rectifier. The rectifier takes the input voltage and changes it from ac voltage (alternating current coming from the utility power) to dc voltage to charge the battery and provide dc power to the inverter. In most online UPSs, the rectifier or filtering is used to make sure that the load, as well as the UPS, does not allow unnecessary noise and harmonics to be fed back into a building's power. Because an online UPS is always creating its own power from the rectifier and inverter, it never has transfer interruptions and can provide a much higher level of power quality.

Keeping in mind that reliability is the single biggest consideration in choosing a UPS, it is possible to enhance the reliability of all UPS topologies by using redundant configurations or UPSs to back up other UPSs. While all the topologies may differ in price and performance, the key to successful power protection is choosing the UPS that is right for your needs, no less and no more.