Doug McCann 2016-03-01 11:21:51
Online Uninterruptible Power Systems (UPS) with capacities of 5 to 150 kVA have been available for almost 40 years, and the market continues to grow at a 12% annual rate as businesses and the industry become more dependent on the continuous flow of clean, reliable power. Various UPS industry marketing surveys reveal that 8 out of 10 UPS systems produced today are designed specifically for IT applications. It’s certainly not surprising that industrial UPS specifiers have a difficult time identifying the critical differences, including cost and long-term support issues between commercial and industrial UPS equipment. To clarify this dilemma, we have identified three major market segments for the application of online UPS systems. They are: Information Technology (IT). The term Commercial UPS has become associated with IT because the UPS and computer room are often purchased as a package from construction contractors. Data centers in banks, hospitals, and insurance companies are examples of typical IT UPS applications. The bidding specifications for IT UPS equipment are not technically demanding and typically require only safety certification such as UL or CSA. The interruption of alternating current (AC) power may disrupt data processing and telecommunications, but does not create an inherent risk of injury to people or property. Critical Process Control. Over the years, UPS systems designed for Critical Process Control applications have earned the name Industrial UPS. Petrochemical complexes and electrical power generation plants, for example, use chemical or steam processes that can become dangerous when AC power is interrupted, causing possible chemical process instability or costly damage. To prevent such risks, UPS systems intended for industrial applications must be designed and performance-tested to a more rigorous level than commercial equipment. General Process Control. There are applications, such as the pharmaceutical and food and beverage industries, that fall into a middle ground that we will label as “light industrial.” These types of applications have processes that are not inherently dangerous, even if AC power is interrupted during process operation. Some of the major application-driven differences between the IT UPS (commercial) and process control UPS (industrial) are summarized in Table 1. A typical IT data center is represented in Figure 1. We can see from the diagram that IT data centers are electromagnetic and radio frequency interference (EMI/RFI) surge-protected environments. The design of the data center’s AC power distribution system protects the UPS from the effects of incoming EMI/RFI on the AC mains and bypass input feeders. Using dedicated input power feeders with the isolation transformers reduces both the EMI/RFI and the available fault current at the UPS AC mains. Competitive pricing pressure and the protected IT electrical environment sometimes justify the omission of the input isolation transformer in the rectifier section. It is crucial that industrial UPS specifying engineers carefully evaluate the UPS electrical environment, as, almost always, a rectifier input isolation transformer is required. Figure 2 shows the industrial UPS rectifier section with an input isolation transformer. In some industrial applications, the UPS AC input power is fed from switchgear or Motor Control Centers (MCC) and often shares bus connections with electrically noisy loads such as variable speed drives. An input isolation transformer may be necessary to protect the UPS from the effects of power feeder surges and RFI/EMI. Environmental Considerations UPS environmental considerations are often driven by the application. In power generation applications in particular, there are typically higher ambient temperatures, (e.g., > 30°C [86°F]) and particulate contamination of the air. Industrial UPS equipment is designed to tolerate moderate amounts of non-conductive dust and high ambient air temperatures of at least 40°C (104°F), and usually has an option for 50°C (122°F) ambient temperature. Geothermal power plants often produce as a process byproduct sulfur dioxide gas, which forms dilute sulfuric acid when combined with moist air. In such cases, industrial UPS systems will have epoxy-coated or nickel-plated copper bus bars, special anti-corrosion terminal connections, and acid-resistant metal surface coatings. In extreme cases, the cooling air to the UPS cabinets is brought in from a clean (or scrubbed) source at a positive pressure and exhausted out of the cabinet tops to keep atmospheric pollutants out of the UPS interior spaces. In contrast, commercial UPS environments are almost always temperature controlled at 30°C (86°F) and kept very clean. The service life of any UPS system is reduced by operation in high ambient temperatures. The critical UPS component most affected by high ambient temperatures is the UPS battery, but other internal UPS components, such as direct current (DC) bus filtering capacitors, may have their service life shortened by high-ambient temperatures unless special high-temp components are selected. UPS service life, ideally 20 years or more, is most important in critical process control applications. That is why industrial UPS systems have built-in design margins to preserve the operation life in tougher environments. In addition, the equipment will also have predictive parts replacement programs to insure high UPS mean time between failure (MTBF). Hybrid vs. Full Electronic Static Switch Types Figures 3A and 3B show the placement and construction of a hybrid or “wrap-around” static switch commonly used on IT UPS systems. Only the bypass pole of the static switch has the inverse-parallel SCR (Silicon Controlled Rectifier) pair for electronic power switching. The inverter side of the hybrid static switch uses a power relay with normally open contacts to disconnect the inverter from the bypass during the normal static switch critical load transfer operation. Contactors are not as reliable as SCR devices, especially if the inverter is exposed to repeated load faults. Under critical load fault, downstream of the UPS output, the short circuit current is limited only by the impedance of the bypass source. The high fault current levels can weld the contactor’s electrical contacts closed. If the UPS hybrid static switch contactor contacts fail to open, the UPS output will be connected continuously to the bypass source. UPS inverters, especially some PWM inverters, are not designed to have their AC outputs backfed from the bypass for more than 30–50 milliseconds. Industrial UPS systems use inverse-parallel SCR devices on both the inverter and bypass poles. This modification to the hybrid static switch is a more expensive design feature, but the elimination of the contactor increases the overall static switch reliability. The MTBF of the UPS is directly linked to the reliability of the static switch, which is in the power path between the inverter and critical load. Using SCR devices on both poles of the static switch creates a more reliable power path by eliminating the inverter-side electromechanical power contactor. An additional benefit is the switching flexibility added by the bypass-side SCR devices. Some Industrial UPS systems have two static switch transfer modes, depending on the state of inverter and bypass phase synchronization (Figure 4). If the bypass and inverter are in phase synchronization, the static switch performs an overlapping (make-before-break) transfer. If the bypass and inverter are not phase synchronized, and the UPS load exceeds the inverter capacity, the static switch will perform a 0.25 cycle break-transfer to prevent the creation of out-of-sync circulating current. This type of dual mode static switch switching flexibility would be impossible without a fully electronic static switch. UPS Contingency Design Analysis Contingency design analysis is the examination of how the UPS behaves when the unexpected happens. In an IT UPS system, for instance, if the internal power supply to the static switch control board fails, the static switch without a true fail-safe design can’t transfer the critical UPS load—the SCR gate drive no longer has the power necessary to operate correctly. Few IT UPS systems use fail-safe static switch designs. Very commonly the SCR gating circuits in the hybrid static switch are triggered by opto-couplers, which require external power to develop the SCR gate trigger pulses. In a fail-safe static switch design, the static switch SCR devices derive their gating power from the load current. A power supply failure will force the UPS to transfer the critical load from the inverter to the bypass. It is common these days for UPS equipment to use microprocessor control. Unfortunately, unless careful thought is given as to how to limit the scope of microprocessor control, single point failure mode creates a problem if the microprocessor fails. In UPS designs that have not considered single point failure modes, the failure of the microprocessor results not only in the sudden loss of UPS output, but also the failure of the static switch to transfer of the critical load to the bypass. In an industrial UPS design, single-point failure modes are carefully considered and eliminated if at all possible. In a well-designed, microprocessor-controlled UPS, both internal and external watchdog circuits monitor the microprocessor’s calculations continuously. If a microprocessor error is detected, the industrial UPS, with its independent and fail-safe static switch, immediately transfers the critical UPS load to the bypass. The Manual Bypass Switch (MBS) gives the maintenance personnel the ability to place the critical UPS load on the bypass source intentionally for UPS service. In an IT UPS, the manual bypass function is usually performed with a three circuit breaker arrangement. The load-to-bypass transfer is first accomplished via the static switch and then sealed or jumpered by closing the manual bypass breakers. While this seems like a simple bypass method, it depends entirely on the proper operation of the UPS static switch. If the static switch is not operational, the manual bypass operation via the manual bypass breakers will not allow a closed transition transfer of the critical load. In contrast, the Industrial UPS uses a make-before-break electromechanical switch. With a closed transition manual bypass switch, the critical load switching is completely independent of the static switch operation. In a well-designed industrial UPS system, the manual bypass switch is mounted remote from the UPS cabinet so that total UPS isolation can be provided while the UPS is being serviced. UPS Batteries and Chargers Because conversion efficiency is a primary design criterion in IT UPS, the transistorized PWM inverter bridges are designed with DC link voltages in the 300–400 VDC range. The high DC voltage bridge not only raises the inverter’s DC/AC conversion efficiency, but also reduces the physical size of the inverter bridge transistor heat-sinks. The 300- to 400-V DC Link voltage requires that the IT UPS use many 30 to 40 six-cell, series-connected, valve-regulated batteries. Multi-cell valve-regulated batteries have become the norm for IT-UPS batteries. The industrial UPS will use a much lower DC Link voltage (60-cell, 125 VDC), which has the advantage of using a UPS battery with fewer inter-cell connections. Industrial UPS specifiers, on the other hand, tend to favor the lower DC voltage link because of the increased battery reliability due to fewer inter-cell connections. Additionally, it is more practical to use single cell batteries. Maintenance personnel can easily jumper out a single cell if an imminent cell failure is detected. In general, IT UPS applications use valve-regulated, lead-acid batteries in the 10- to 30-minute ranges. Because the IT UPS battery support times are usually not as long as those used in industrial applications, the IT UPS charger capacity is usually limited in current capacity. They are typically sized to recharge a 15- to 30-minute lead-acid UPS battery to 95% capacity in eight to 10 hours. However, the charger capacity in industrial UPS applications—like power generation using four- to eight-hour UPS batteries—often has to be much larger because the battery support times can range from 60 minutes, to eight hours, or longer. UPS specifiers need to make sure the UPS system being considered has enough battery recharge capacity built into the charger. It is common in IT UPS applications to discharge the lead-acid UPS batteries more deeply, usually to the end of discharge voltage of 1.65 V per cell (VPC). Discharging the UPS batteries to a lower-end voltage will remove more energy from the cells and result in a smaller UPS battery. In contrast, industrial UPS systems with longer battery support times will use a higher-end voltage of 1.75 VPC. Keep in mind that deeper cell discharge reduces the service life of the UPS batteries. Lead-acid cells, especially the less expensive five- to 10-year service life cells, are sensitive to depth of discharge. The UPS Specifier needs to weigh the consequences of reduced battery service life when lead-acid UPS batteries are discharged at less than 1.75 VPC. Equipment Design Life: Matching UPS Service Life to the Critical Process Service Life In the commercial UPS marketplace, UPS models tend to become obsolete in five years, and today’s data center technology typically lasts less time than that. IT UPS suppliers who know their marketplace well focus their design efforts on making next year’s UPS model cheaper, smaller, and more efficient. Rapid IT UPS product obsolescence makes long-term support of older UPS equipment in the field very difficult. By contrast, industrial UPS equipment will have design margins built into its components so that the UPS system will have more than 100,000 hours of MTBF when operated in typical industrial environments. There are components like cooling fans and DC filter capacitors that degrade with time, even with conservative design practices. Industrial UPS users, such as power generation plants, commonly specify UPS service lives in the 20- to 30-year lifespans. Petrochemical UPS specifiers have a shorter time horizon, usually in the 10- to 15-year range. In general, however, industrial applications, in which the process drives the decision, need a longer UPS service life. Industrial UPS suppliers will have documented component replacement schedules—components like cooling fans and DC capacitors—so that the UPS MTBF can be maintained over the 20- to 30-year service life. Industrial UPS suppliers will keep UPS designs and UPS spare parts in production for longer intervals because they are geared to support the older UPS systems in the field. Conclusion Significant differences do exist between commercial and industrial UPS applications and equipment. When selecting a UPS supplier, it is important to consider not only the application for the system, but also the long-term support. When one considers the life-cycle costs of a UPS—especially in large process applications—an industrial UPS design, with a scheduled critical parts replacement program, is a lower-cost solution than UPS replacement. BE Doug McCann is an Industrial UPS Consultant with Ametek Solidstate Controls.
Published by Forester Media. View All Articles.