Most of the time you don’t need your UPSs, so they are “out of sight and out of mind”. Meanwhile, the batteries which provide the energy when the power goes out, are very sensitive to the environment.
UPS systems: Threats, maintenance, management and winning designs
By Jack Ward, MD of Powermode
Despite Eskom’s pronouncement that ‘loads-shedding is a thing of the past’ most electricity users continue to experience regular power outages. Now under the banner of ‘grid maintenance’, these outages continue to plague businesses.
In order for a standby system – such as a battery-powered uninterruptible power supply (UPS) system – to take over the load on demand, the importance of regular inspections, service and maintenance regimes cannot be overstated.
UPS systems are constantly on standby, powering silently in the background. Imperceptibly, they face a number of threats to their wellbeing on a daily basis.
If a UPS system fails it is almost certainly to be at an inopportune moment – when a power outage occurs. In such instances, the computer system the malfunctioning UPS supports will shut down and valuable data could be corrupted or lost. The cost to the company in terms of downtime will be significant.
Probably the number one cause of UPS malfunctions is battery failure. Battery life is influenced by many factors including storage conditions, ambient temperature, battery chemistry issues and shelf-life. As a rule of thumb, a UPS battery has a life cycle of between three and six years
The key to battery longevity lies in a thorough understanding of the status of individual batteries in a multi-battery pack in terms of their duty cycles and load factors.
Battery maintenance should be top of a UPS user’s priority list. A good maintenance regimen will help prolong battery life while keeping a note of where the battery is in its life cycle will provide an indication as to when failure is imminent.
The prevalence of unexpected power outages in SA – often longer than four hours – has exposed one of the Achilles’ heels of standby power devices; the dramatically shortened lifespan of batteries (including deep-cycle batteries) when subjected to full depletion on a regular basis.
To help prevent premature battery failure, a new-generation, battery management harness is a necessity. The ‘smart’ battery technology in leading harness brands monitors and optimises the performance and efficiency of individual batteries in a battery pack on a 24×7 basis.
This is achieved by a dedicated, computerised battery balancing solution that automatically monitors data streams containing information critical to the well-being of individual batteries, including temperature, state of charge and depth of discharge. The tally of the number of discharge/ charge cycles is recorded.
Using this data, the system’s battery performance is able to be balanced and equalised. Should an individual battery’s operating parameters not meet design specifications or fail for any reason the battery is flagged for replacement – thus ensuring the integrity of the pack.
The technology featured in a battery management harness not only distributes and balances the battery load, but discharge and charge regimes across all batteries in the pack can be monitored on a minute-by-minute basis. Thus steps can be timeously taken to ensure that no battery is compromised through over-cycling or a malfunction of any kind.
Apart from battery failure, other serious threats to the smooth operation of a UPS system include a sudden ‘spike’ in the power supply. Called transient spikes, they are likely to inflict serious damage to the input side of a UPS – the filter/rectifier siting.
Regular maintenance regimens will most certainly be able to determine whether damage has occurred and remedial steps can be initiated. This applies to almost all elements of a UPS system,
The scourge of Highveld weather, lightning, can do serious damage to a UPS system and to the highly sensitive computer systems it is tasked to protect. A common misconception is that a UPS system constantly protects itself and the equipment load from lightning strikes. If the amount of energy in the transient (the lightning strike) is large enough, damage will occur. Surge suppression devices, if installed, including metal oxide varistor (MOV) devices are often compromised by lightning strikes.
As small as a two-rand coin or as large as a cool-drink can, capacitors, like batteries, degrade over time. The effects of time many not be apparent, but a single failure will have a domino effect, leaving the other capacitors to work harder and fail sooner. Most UPS systems contain as many as a dozen or more of these simple devices that store and release electrical energy.
Dust is the enemy, steadily advancing to block filters and cause progressive overheating of UPS systems. Regular monthly inspections are required to address this problem. Fortunately, filters are the least most expensive components of an effective UPS maintenance plan.
Contactor failure is another threat to UPS systems. Contactors, which are also prime collectors of fine dust and other resistive particles, require regular inspection and cleaning to ensure optimum performance and to guard against premature failures.
The failure-proofing of relays is not usually at the top of any maintenance schedule, however, technically-aware UPS owners understand that the sticking of welded relays may go unnoticed for long periods of time. The problem is revealed only when emergency change-of-state events occur. Appropriate inspection procedures are able to detect problems before they arise.
Total cost of ownership
Selecting a UPS system that matches an organisation’s exact needs is vital to UPS longevity. Consulting a knowledgeable provider with experience in the UPS industry before making an investment is prudent as is selecting a system that can organically grow in tandem with the company’s expansion plans.
In this light, when budgeting for a UPS system, it’s crucial to consider the total cost of ownership (TCO) – rather than just the original purchase price. When operating costs and upgrades are taken into account, a unit with an apparently cheaper initial purchase price can often prove to be more expensive in the long run than a better-technology solution at a slightly higher original price point.
In reality, a UPS featuring modern, modular technology can reduce the TCO significantly, while improving reliability and dependability.
To secure first class UPS protection, business managers should accept that the best-performing UPS technology – as found in the latest modular systems – will cost a little more, but within a year the cost difference could well be recovered in its entirety.
Before explaining how this is possible, the evolutionary path along which UPS technology has travelled need to be examined.
Legacy UPS systems, featuring transformers, were large and heavy when compared to today’s systems. For example, a data centre with a 120 kVA load could theoretically have been supplied by a single, cumbersome, floor-standing 120 kVA unit.
However, because fail-safe redundancy is a likely requirement to ensure availability, this would demand the fitment of two 120 kVA units sharing the load in a 1+1 redundant configuration.
For the organisation that the two UPSs’ served, it meant investing in substantially more capacity than actually necessary. It also meant that neither UPS unit could ever be more than 50% loaded, which for a transformer-based system results in a significant reduction in efficiency.
With the advent of transformerless technology has come much smaller and lighter UPS solutions which can easily be incrementally added to a racking frame to achieve an application’s required power capacity and redundancy targets. Unwieldy, free-standing individual units are now the dinosaurs of the standby power world.
In a hypothetical scenario, the 120 kVA load can now be met by a single rack containing four, 40 kVA ‘hot swap’ plug-in modules. The load remains fully supported with n+1 redundancy, while the total UPS capacity has been reduced from 240 kVA to 160 kVA.
Although the purchase price per kVA for modular UPSs will be slightly higher than for legacy types, this difference will be partly offset by the reduction in purchased capacity – and in the floor space required for installation.
Moreover, significant savings in operating costs will also be made as the modular solution is more efficient than a transformerless implementation – especially one that cannot operate at more than 50% loading. Considering the 120kVA example as discussed, over a five-year period, savings could be as high as R 500,000.
And that’s not the end of the story. A modern modular UPS systems’ can slash operating costs even further by reducing the need to hold emergency spare parts. Instead of a slew of costly spares that might be needed, a single spare plug-in module will suffice.
This is true even when modules of different power ratings are being used, because simply holding a module of the highest kVA rating installed will cover all eventualities. A trained technician can hot swap a UPS module in under 5 minutes.
Modular system upgrading is also far simpler, faster and cheaper as extra capacity can be added simply by plugging in additional modules without even interrupting power to the critical load. The lengthy building work, sizable increase in footprint and frustrating interruption to supply associated with extending traditional systems is completely eliminated.
One final point: A UPS system’s availability is increased if its mean time to repair (MTTR) is reduced. An attractive feature of a modular UPS system is its almost zero MTTR figure. If a hot-swappable module does fail, it can be withdrawn from the UPS frame without interrupting power to the load. A replacement module can be plugged into the rack immediately.
By contrast, if a legacy system fails, it must be shut down, isolated from its mains supply and repaired in situ; a process that typically takes five or six hours to complete. This means that, unlike a legacy transformer-based system, a modular UPS solution can provide ‘six-nines’ (99.9999%) availability which equates to just 5 mins of downtime per year.