Cutting back on preventive maintenance procedures you’re your uninterruptible power supply (UPS) systems during tough economic times can spell disaster down the line
Considering the current economic climate, companies of all shapes and sizes are looking to cut costs any way they can, including delaying the purchase of new equipment and deferring maintenance (PM) activities. Although this mindset is understandable, it's important to remember that postponing preventive maintenance can end up costing you much more in the long run.
When it comes to uninterruptible power supply (UPS) systems, for example, minimal or no maintenance greatly increases the chance that business operations will be disrupted if the power equipment fails, thus exposing the company to loss of revenue, reducing work productivity, affecting customer satisfaction and loyalty, etc. That's not to mention the costs that can be incurred for repairs and replacements. So in reality, PM on existing equipment is more important than ever during tough economic times.
Benefits of PM
When equipment is not properly maintained — especially under adverse conditions, such as dirty environments and/or high temperatures — system deterioration and load loss can result.
PM programs maximize the reliability and performance of UPS systems on which organizations depend to keep critical systems running. When correctly implemented, PM activities ensure maximum reliability of data center equipment by providing systematic inspections, detection, and correction of incipient failures — either before they occur or before they develop into major defects that could translate into costly downtime later on. Typical PM programs include inspections, tests, measurements, adjustments, parts replacement, and housekeeping practices.
Frequency of maintenance
How often you schedule PM procedures depends on the type of UPS. Small UPS devices should be inspected annually to ensure alarms, filters, and internal batteries are operating within specifications. For medium and large systems, which most likely include ancillary equipment, inspection and maintenance should take place at least twice a year.
Every UPS is designed with components that have a limited life — some of which live longer than others — depending on the design of the component itself, its function within the UPS, and the operating conditions to which the UPS is subjected. Even though these key components have a limited design life, a well-planned maintenance service, which includes the periodic replacement of components, can ensure continuous and reliable operation of the UPS for many years.
Start at the component level
A UPS has a design life well in excess of 10 yr. In fact, well-maintained units can continue to provide economic benefits for 20 yr or more. Long-life components are used in a UPS whenever it is practical and cost-effective. However, some components are more sensitive than others; therefore, their exposure to temperature and voltage fluctuations can wear them out prematurely. The typical limited life components found in a UPS are listed in Table 1 (click here to see Table 1), as well as the design life and time at which replacement is recommended.
Generally, power capacitors are considered “failed” when their measured capacitance is 5% below the nameplate rating. Typically, storage batteries are “worn out” when their full load reserve time is 20% below rated. Site-specific battery plants may have different end-of-life requirements. Fans and blowers typically will stop rotating or rotate slower than expected.
In most cases, replacement components must match the original component specifications exactly. Unfortunately, many are not readily available from third-party component distributors.
The reliability of any system is only as good as the shortest component life in the unit. To prevent a failure of any of the parts affecting the critical load operations, periodic inspections and replacements are recommended before they reach their expected wear-out life. Following are some guidelines for specific components.
Transformers, inductors, and DC chokes. The design value of most magnetic components is 40 yr of operation. Key factors affecting these components are the insulation used in the winding process and the temperature rise while in service. The best magnetics use Class H insulation rated at 220°C and are never operated at higher than 150°C.
Some commutation inductors (used in SCR-based UPS modules) may exhibit deterioration after 10 to 15 yr of service due to load variations and the constant cycling of the current through the inductor. Before it fails, the inductor starts to vibrate and becomes very noisy. This condition will be obvious during PM.
Power semiconductors. These devices do not have a rated end of life in the normal mode of operation of the UPS. SCR or IGBT failures are generally secondary symptoms of other problems.
During the annual PM cycle, semiconductor devices should be visually inspected for corrosion and damage to the hermetic seal. If corrosion or seal damage is found, the device should be replaced. In addition, gate leads should be inspected and mounting torque of the semiconductor hardware checked. Changing the power semiconductors in a large UPS at regular intervals is not recommended.
Electrolytic DC capacitors. The expected life of electrolytic capacitors can be calculated as a function of the manufacturer's rating and the expected operating temperature of the device. Please refer to Design Life Calculations for Electrolytic Capacitors. The theoretical service life of DC capacitors ranges from eight to 30 yr, depending on system bus voltage and UPS ambient temperature.
Experience has shown that capacitors should be replaced well before the end of their rated service life. This helps account for variations in ambient temperature, airborne contaminants and other environmental factors. We recommend replacing DC capacitors every six to seven yr.
Oil-filled AC capacitors. Oil-filled capacitors have a design life of 10 yr, and a replacement life of seven yr. As oil-filled capacitors approach their design life, they are subject to the internal breakdown of their soggy foils and possible loss of capacitance. It is recommended that all oil-filled capacitors be inspected, and those within six months of their replacement life (seven yr) be changed out during the annual PM cycle. These capacitors should also be inspected during the annual PM for deformation, which indicates those that need replacing.
Circuit boards. There is no rated service life on the components specified for use on circuit boards. Problematic circuit boards can be removed and returned to the manufacturer for repair and testing. Before they are returned to service, they will have all outstanding revisions incorporated and will then be system tested. If a circuit board fails a second time for the same problem, it is scrapped. All calibrations are verified during the annual PM process to ensure the circuit boards don't exhibit signs of failure. If weakness (any sort of defect in the circuit board) is found during the PM process, the circuit board should be replaced.
The most serious limitation to circuit board longevity is availability of replacement components. Certain parts are no longer available from manufacturers. Although manufacturers carry safety stock on some key components, it's difficult to foresee all contingencies. This parts-availability issue affects all vendors of static and rotary UPS products.
Batteries. Battery maintenance begins with installation of your system. Batteries must be fully charged, battery room conditions verified, and baseline ohmic readings recorded for proper trend analysis throughout the life of the battery. If this information is not properly gathered and documented, determining bad batteries could prove difficult.
For best practices for battery maintenance, refer to the manufacturer's recommendations, IEEE Std-1188 for Valve Regulated Lead Acid (VRLA) batteries and IEEE Std-450 for Vented Lead Acid (VLA or flooded) batteries. However, it's important to remember that best practices do not always equate to common practices. Governed by real-world factors, many facility managers are often forced to take into account the cost of performing the recommended schedule as it relates to the criticality of the application.
Table 2 (click here to see Table 2) represents a typical PM schedule for VRLA and VLA (flooded) batteries.
High ambient temperature and frequent discharges are most commonly responsible for reducing useful life across all types of batteries. (Dryout is the most common cause of VRLA battery failure.) Battery aging accelerates dramatically as ambient temperature increases. This is true of batteries in service and in storage. Even under specified temperatures, batteries are designed to provide a limited number of discharge cycles during their expected life. Although that number may be adequate in some applications, there are instances where a battery can wear out prematurely.
An ounce of prevention.
All UPS models are designed and manufactured with components that have limited life. The reliability and continuous operation of a UPS can significantly be improved over the years by properly maintaining it.
This PM process must include a periodic system inspection and replacement of key components that, by design, have a limited life. Some of these components may last longer than others, but the reliability of the system can be compromised by the component with the shortest design life. Therefore, it is strongly recommend that any UPS system be periodically inspected and key components be replaced to ensure optimum support to the customer's critical load.
Hu is a Liebert service product manager, power and Donato is a Liebert service product manager, batteries with Emerson Network Power in Westerville, Ohio. They can be reached at firstname.lastname@example.org and email@example.com.
Sidebar: Design Life Calculations for Electrolytic Capacitors
The expected life of these capacitors has been confirmed to follow Arrhenius' equation — a formula describing chemical reactions due to dielectric molecules activated by thermal energy. You can calculate L, expected device life in hours, according to the following:
L 5 LBASE 3 2(TBASE2TCORE)410 3 voltage multiplier
LBASE 5 5,000 hr at 85°C core temperature and rated voltage
TBASE 5 85°C
TAMB 5 30°C
TCORE 5 TAMB1 TRISE-AVG 5 30°C 1 10°C 5 40°C
VR 5 rated voltage
VA 5 actual voltage
Voltage multiplier 5 (2.50 2 1.5) VR4VA
5 700 4 540 5 1.2963 for 540VDC bus (480VDC nominal)
5 500 4 405 5 1.23 for 405VDC bus (360VDC nominal)
Solving this, we get:
L 5 5,000 3 2(85240) 410 3 1.2963
5 5,000 3 24.5 3 1.2963
5 146,659 hr (16.7 yr) for a 540VDC bus and 30°C
L 5 5,000 3 2(85240)410 3 1.23
5 5,000 3 24.5 3 1.23
5 139,158 hr (15.9 yr) for a 405VDC bus and 30°C
Substituting 40°C for UPS ambient temperature, we obtain:
L 5 73,329 hr (8.4 yr) at 540VDC and 40°C
L 5 69,838 hr (8.0 yr) at 405VDC and 40°C