Keep on running

Electrical power is perhaps the most essential raw material used by commerce and industry today.

It is an unusual commodity because it cannot be stored conveniently in bulk and cannot be subject to quality assurance checks before it is used. Despite this, many businesses are increasingly reliant on high quality and secure power supplies for data processing and industrial applications - making power quality a key issue in electrical design.

The perfect power supply would be one that is always available, within voltage and frequency tolerances and has a pure ‘noise-free' sinusoidal waveform.

Power quality defects include blackouts, under- or over-voltage, dips (or sags) and surges, transients, distortion of the supply as a result of harmonic effects and frequency variation. Just how much deviation from perfection can be tolerated depends on the user's application, the type of equipment used and the user's view of their requirements.

The UK power supply industry regularly claims a supply availability of 99.98%. This sounds impressive, but to the average customer it represents unpredictable disconnections totalling 97 minutes per year, with 90% of all customers experiencing interruptions of significant duration.

These figures are also misleading because interruptions of less than one minute are often not reported. For many commercial operations, an interruption of one second is just as disruptive as an interruption of, say, 10 minutes. Reports indicate £200m was paid out by insurers for losses associated with power cuts in 1995 (the latest figures available) and it is likely that this figure is much higher now.

Power quality solutions

To overcome power quality problems, the following solutions may be employed:

Duplicated incoming supplies.

These are derived from separate sub stations and follow different routes to the site. However, the cost of providing such an arrangement can be prohibitively expensive and changeover from one supply to another is normally a manual operation by trained staff, which can take several hours.

Duplicated power distribution.

Diversely routed power distribution within a facility will allow it to continue to operate if there is a failure in an LV panel and its cabling.

Standby generators.

On detecting a mains failure, the standby generator will cut in and supply the load once the generator has run up to speed, so that it is ready to accept the load (typically in around 30 s). It should be noted that unless the standby generator is rated to supply the full site load in a single load step, the shedding of non-essential loads is required to ensure it does not become overloaded. Unless the standby generator is approved to work in parallel with the mains supply, there will also be a power interruption upon restoration of the mains supply when load is transferred from the standby generator back to the mains.

Uninterruptible power supplies.

These are often seen as an essential feature where business-critical equipment is in use. UPS installations are designed to continue supporting the load during a failure of the mains supply without interruption. It should be noted, though, that UPS equipment can introduce undesirable effects on to the power supply - so that by protecting against power failures, power quality can be compromised.

Surge suppression.

Devices can be incorporated at major wiring centres to provide protection against surges and spikes in the supply as a result of events such as lightning strikes or switching of major loads on the utility supply.

System configuration.

Major loads should be powered directly from the main distribution to avoid switching transients and voltage drops affecting other parts of the system.

Reliability/resilience considerations

To decide on an appropriate power protection strategy, it is necessary to undertake a quantitative reliability study taking account of system and component Mean Time Between Failure (MTBF) and Mean Time To Repair (MTTR).

It is also necessary to consider resilience of the system to failure. N+1 is a common equipment arrangement where, if N items of plant are required to meet the load, N+1 units are provided to ensure there is a spare item of plant available to cover for failure of any of the N items. N+N is a higher standard, giving a truly redundant configuration without any common points of distribution.

The system should be analysed for single points of failure that could negate other measures built in to the system to improve reliability. The electrical system should be analysed both with and without mains being available to ensure compliance with wiring regulations under both scenarios, as this is normally harder to achieve when operating with a standby generator.

The Uptime Institute publishes ‘Tier' definitions to categorise system configuration, expected availability and presence of single points of failure. Tier ratings run from I-IV, with Tier IV currently being the highest standard which is based on double redundancy with a full System + System arrangement, no single points of failure and an expected availability of 99.995%.

Statistically, this means that a Tier IV facility is likely to be unavailable for around 26 seconds per year, reflecting the probability of both systems failing simultaneously or human error, which is recognised as a significant risk. These definitions are shown in Table 1, above.

Safety critical industries such as aviation and nuclear often require higher levels of reliability - the ultimate being accepted as five nines, or 99.99999% reliability. This is normally achieved by employing triple redundancy and addressing human factors that can contribute to failure. Systems should be subject to an integrated systems test at handover to demonstrate that they operate as designed under all failure modes and regular testing should be undertaken to ensure that reliability is maintained.

Harmonic distortion

Harmonic distortion is a measure of the amount of power contained in the harmonics of a fundamental signal. Harmonic frequencies are integral multiples of the fundamental. For example, for a fundamental frequency of 50 Hz, the second harmonic would be 100 Hz and the third would be 150 Hz.

Most cyclical waveforms can be deconstructed to a sinusoid at the fundamental frequency plus a number of sinusoids at harmonic frequencies. Figure 1, left, shows a typical waveform that has been distorted by harmonic effects and the various terms used to describe the distortion.

All ‘non-linear' loads generate harmonic currents. These include: Switched Mode Power Supplies (SMPS) found in most computers and electronic equipment, fluorescent and high frequency lighting ballasts, variable speed drives, UPS and magnetic cored devices.

A series of studies by BSRIA showed that eight out of 12 sites it investigated experienced significant problems associated with harmonic distortion. These

included high neutral currents and elevated earth voltages, leading to currents of up to 48 amps.

But it is possible to design and build electronic equipment that does not introduce harmonics. Legislation has been brought in recently to limit the harmonics created by SMPS equipment. However, this has resulted in equipment being fitted with capacitors that result in unity or leading power factors, which can create other problems.

Designers should select equipment that reduces harmonic distortion by specifying higher order (12 or 24 pulse) rectifiers, as opposed to 6-pulse equipment. Harmonic distortion can be significantly reduced and system efficiency improved.

UPS

Uninterruptible power supplies can provide a degree of isolation between the power source and load and can isolate the user from transients and interruptions in the main supply.

However, problems arising from equipment supplied by the UPS will still be present. And since the supply impedance of a UPS is usually much higher than the mains supply, voltage distortion in systems supplied by UPS equipment can be much worse.

When a number of loads are connected to the same supply, differences in phase angle will produce a cancellation effect that can reduce the total harmonic distortion within the system. Careful design of the system can lead to a reduction of harmonic problems without the need for artificial measures being implemented.

Earthing

There is an increasing trend toward using the earth of the building as a reference for electronic equipment, which leads to leakage currents at the fundamental frequency.

The primary purpose of the earthing system for a building is to ensure that a safe environment is maintained and to protect equipment from damage in the event of a fault. This is achieved by providing a low resistance path for earth fault currents so that protection equipment can operate quickly; by establishing an equi-potential platform on which equipment can be safely operated; and by bonding exposed metalwork to earth.

The high density of electronic equipment in modern commercial buildings means that earth leakage currents can be significant. Because the earth system is required to carry currents, any break in the conductor will mean that the isolated section will rise to a potentially dangerous voltage. For this reason, the wiring regulations require that in these situations a high integrity earth be provided.

For areas having high concentrations of electronic equipment, there is often a requirement to provide a ‘clean' earth in addition to the protective earth. This has commonly been provided in the form of a star-connected clean earth running directly back to the main earth terminal. It should be noted that such earths are often subjected to significant currents/voltages and noise so that the so-called ‘clean' earth is often far from clean, compared to its ‘dirty' counterpart.

The star-connected earth can be subjected to resonance effects when the conductor lengths are long and the length is close to certain fractions of the wavelength of the noise frequency, depending on the relative values of the conductor reactance and capacitance.

Also, it is extremely difficult in practice to separate, and maintain separation of, clean and dirty earth paths and this can lead to currents flowing in loops formed by ground wires of data cables.

This problem can be overcome by the use of a mesh earth arrangement, whereby the raised access floor acts as a continuous ground reference plane with multiple earth paths back to the main earthing terminal.

Identifying problems

Predictive studies are not always undertaken for new facilities. In reality, they may only be necessary for facilities housing large amounts of IT equipment, high frequency lighting, inverter drives etc that can generate harmonic currents that need to demonstrate compliance with engineering recommendation G5/4.

The photograph above shows a current measurement being taken by two separate meters on the same cable. One device shows a current of 59.2 A being drawn and the other 40.5 A, because the meter with the lower reading is insensitive to higher frequencies. This illustrates the importance of using the correct equipment when undertaking harmonic studies.

Harmonic analysers are available to investigate harmonic distortion throughout the distribution network. Prior to undertaking this type of work, reference should be made to Regulation 14 of the Electricity at Work Regulations - a key safety standard when working on or near live apparatus.