Measuring Power System Performance for High-Reliability Applications

Power system performance isn’t about uptime “nines.”

Even if your facility doesn’t incur the $10 million/min cost of downtime that some facilities experience, unplanned downtime is costly. Power quality plays a central role in the downtime drama, but focusing on the availability of power can cause you to overlook it in budgeting, planning, and practice. The truth is, paying more attention to power quality issues than the bottom line “nines” count will improve system performance and reliability.

Measuring power availability won’t help identify voltage sags, transients, voltage unbalance, voltage regulation concerns, or harmonic distortion levels, all of which could be important to the facility or the design of the power conditioning equipment. A systematic approach to monitoring takes all of these factors into account. To develop such an approach, you first must define your objectives and how they apply to your facility.

A lot is expected of monitoring systems at high-reliability facilities. Among other things, they must:

Continuously evaluate the supply for disturbances and power quality variations that could degrade or disrupt operations.
Document the performance of your power conditioning equip- ment, such as static switches, UPS systems, and generators, during normal system conditions and during system disturbances.
Evaluate power quality characteristics of equipment, including harmonic interaction between loads and power conditioning equipment; inrush characteristics for loads, such as compressor motors that can affect back-up generator operation; transients associated with switching events; or the response of equipment to voltage sags or transients from the power system.
Document and describe disturbances that disrupt your facility’s operation.
Document energy use for each part of the facility, which helps allocate costs and improve subsequent facility designs.

Proactive monitoring. The traditional approach to power quality monitoring is reactive. People install monitoring equipment after a problem occurs with the idea that they’ll be able to “see” it the next time it occurs.

Permanent power quality monitoring systems help identify conditions and events that may cause problems. Characterizing harmonic distortion levels as the facility loads grow can help identify conditions that may be a problem for facility wiring, the UPS system, or backup generators. Identification of disturbance characteristics associated with motor starting or other load switching may be critical in evaluating equipment ratings and protection requirements for facility loads. Changes in characteristics like voltage unbalance and regulation, harmonics, or transients can indicate an emerging problem.

If a problem does occur, you need to know the conditions associated with it. What was the supply from the utility during the problem? Was there an interruption, a sag, or a capacitor-switching transient? What was the response of the power conditioning equipment? Did switching equipment operate correctly? What was the response of the load? Did the load and the system interact?

Power quality variations. Your power quality monitoring system should be able to monitor a full range of performance characteristics and display useful information in a convenient form in real-time. The system must characterize steady-state power quality irregularities, such as voltage regulation, unbalance, and harmonics, by using time trends and statistical distributions. It also must characterize disturbances like transients, voltage sags, interruptions, and outages with detailed information about the events and provide trending. The Table below summarizes characteristics of power quality variations. Let’s look at some examples of power quality variations.

Voltage sags trip chillers. Chillers—usually unprotected by a UPS—are subject to power quality variations in the supply. A momentary voltage sag could cause the compressor motor to overheat and trip out, but the monitoring system can help you troubleshoot, and correct for, such an event. Corrective action could include selectively shutting down chillers to prevent overheating while limiting operations until the facility can ride through the sag.

Transients trip adjustable speed drives. The advantages of adjustable speed drives make them almost mandatory, but voltage sags and capacitor-switching transients on the supply system can cause them to trip. Characterizing performance during these disturbances can help you optimize the drive controls and settings to avoid problems.

Motor restarts after a disturbance cause generator problems. Motor loads typically drop off during an interruption. Because of process demands, they must restart quickly once you have power from the generators. System cooling in data centers is important—the chillers must restart as soon as possible. Monitors should characterize the voltages at the main supply and at the backup generator, so you can correctly assess the performance of the overall system.

Harmonic loading affects facility design requirements. The design of the facility wiring and the size requirements for UPS systems or other power conditioning equipment depend on the load characteristics, including the harmonic content. High third harmonics, which are typical of computer loads, call for higher-rated neutral conductors and additional capacity or filtering in the power conditioning equipment. The monitoring system should characterize the harmonic distortion levels and help evaluate the requirements for wiring, power conditioning, and transformers based on these characteristics.

Monitoring locations. The selection of monitoring locations depends on the facility design, critical loads, power conditioning equipment, and the specific objectives of the monitoring. At a minimum, monitoring should include the utility supply locations, outputs of power conditioning equipment, and backup generators. If you have redundant or backup supplies, monitor each feed. More extensive monitoring includes critical air conditioning loads and individual loads within the facility, such as communications equipment and individual load buses. Monitoring within the facility can help characterize load interaction issues.

Information from power quality monitoring should be available immediately and conveniently. Today, the best option is to make the system accessible via your intranet or the Internet (Fig. 1 right). These systems connect directly to the network with TCP/IP communications.

Monitoring system implementation. The DHL Air- ways customer service center in Tempe, Ariz., is a good example of a high-reliability facility. This newly constructed 24/7 customer service center serves more than half of the United States and is integral to the company’s customer service functions.

The UPS system consists of three 300kVA parallel redundant units, each containing three 100kVA converter/inverter modules for N11 reliability. Besides protecting the load from disturbances, the design considers the power quality requirements of the supply (low harmonic injection) and the loads (low harmonic distortion output).

As part of the facility design, DHL installed a power monitoring system on the input and output of the UPS system. The system allows access to all power quality monitoring information via a Web browser and automatically sends an e-mail to a list of recipients when it detects a problem or alarm, notifying the appropriate people of the problem and giving DHL worldwide access to the information. This was important because the facility manager travels frequently and needs access from where ever he may be. He wanted to remotely discuss any issues with local people at the facility and see what they were seeing. Fig. 1 illustrates the monitoring system concept implemented by DHL.

Monitoring at the utility supply (Fig. 2 right) indicated more than 50 disturbances in just the first 3 months. These disturbances consisted of sags and transients that could have affected unprotected loads. The system verified the expected performance of the UPS system and failed to detect any disturbances on the UPS output (Fig. 3 below). In fact, the voltage never varied by more than a few percent from nominal. A subsequent investigation found a faulty relay on the utility side.

Power quality monitoring plays an integral role in high-reliability facilities. The monitoring characterizes the performance of the supply system and the performance of power conditioning equipment, including interaction issues with facility loads. Advanced data analysis functions and alarms are making these systems even more valuable, and access via a Web browser allows for teams to collaborate from multiple locations. Good power quality is a journey, not a destination. Plan your trip well—a systematic approach to power quality can make high-reliability a reality.

McGranaghan is a vice president with Electrotek Concepts, Knoxville, Tenn., and Ignall is an applications manager with Dranetz-BMI, Edison, N.J.