Power, interrupted

The recently launched CIBSE Guide K: Electricity in buildings is a useful reference document for those involved with the design, installation and maintenance of electrical systems in buildings.

Statutory and local licensing requirements, plus the functional needs of today’s buildings, have led to the inclusion of a new generation of energy equipment and new standards of reliability. Section 2 of Guide K covers the current legislations relating to electrical systems in buildings.

In many buildings, interruptions in electrical power supplies to vital equipment can cause serious damage, consequential losses and, in some cases, health hazards.

Disturbance and distortion on the 400/230 V supply due to surges and harmonics are becoming commonplace in the electrical installation of almost every building. Their adverse effects on the operation of sensitive electronic equipment are well known to design and installation engineers.

Data corruption and system shutdown are typical consequences of harmonic distortion and power disturbance. Micro-brownouts and long blackouts are examples of events that should be guarded against.

To solve a supply quality problem, one method is to use high tolerance equipment and avoid the use of systems that are sources of disturbances and harmonic generators. This will require the designer to liaise with the end-user to identify then quantify these sources.

This process is not normally plausible in fast track or speculative projects. The duty-holders such as designers, installers and facilities managers must therefore ensure that they can continue to provide electrical power to essential services even in the event of prolonged or short disruption to the mains supply.

Achieving quality power

There are two acceptable techniques to ensure a good quality power supply. Which is adopted will be dependent on meeting the criteria agreed by the designer, installer and facilities engineer, as well as the legislative requirements.

The first method is the use of two (or more) independent mains feeders to supply electricity to the building. Here the main switchboard must be rated for dual-feeder and dual-input configuration. Power is available on both feeders at all times, with one in use and the other on standby. In this application, one has to calculate the availability of the power supply from the two or more independent parallel sources of energy. Availability of the mains supply can be seen as the percentage of time when system is operational. Availability (A) of system can be obtained by the general formula:

A = MTBF

MTBF +MTTR

where:

MTBF = mean time between failure;

MTTR = mean time to repair.

Availability is typically specified in ‘nines’ notation. For example 3-nines availability corresponds to 99·9% availability. A 5-nines availability corresponds to 99·999% availability. Downtime per year is a more intuitive way of understanding the availability. Table 1 (above) compares the availability and corresponding downtime.

There is a simplified calculation method based on downtime. The independent feeders can be considered to be operating in parallel, so the combination is considered failed when all feeders fail. The combined system is considered fully functional if any one of the feeders is available. From this it follows that the combined availability is 1. The combined availability for two feeders can be calculated by the equation below:

A = 1 – [(1–Ax)(1–Ay)]

where:

A = overall availability;

Ax = availability of feeder x;

Ay = availability of feeder y.

The implications of the above equation are that the combined availability of two feeders in parallel is always much higher than that of a single feeder.

Table 2 (p53) shows the availability and downtime for individual and parallel combinations of two or more feeders. From the table it is clear that even though a very low availability feeder was used, the overall availability of the system was much higher. Thus parallel operation provides a very powerful mechanism for making a highly reliable system from low reliability. For this reason, all mission-critical systems are designed with parallel-redundant feeders.

The UPS alternative

The fluctuations, transients and service interruptions common to local utility power sources are generally incompatible with the needs of sensitive critical equipment installations. The second method commonly adopted by engineers is the use of an uninterruptible power supply (UPS) to suppress line transients and provide a reliable and continuous power supply instead of a duplicated feeder supply. Section 9 of CIBSE Guide K details the different types of UPS available and their relative strengths and weaknesses.

In summary, a UPS contains the rectifier and inverter. Both are powered by solid state electronics which are interfaced with programmable controllers. At the output there will be high speed, power-switching circuit-breakers, plus operator command controls and status indicators. During normal operation, a three-phase input is rectified to provide the dc necessary to charge and maintain the storage battery power supply.

The offline UPS is the simplest; typically it has the lowest unit price. If the mains drops below a preset critical value (eg 230 V – 10%), the UPS will draw power from the battery providing a certain degree of protection against power loss or undervoltage to the load.

It does have some advantages: low cost, lightweight and small footprint. However, there are also disadvantages: there is very little output voltage regulation, ie the output voltage can vary between 230 V ±9%. Secondly, the transfer of ac to dc battery inverter supply may take as long as 5-15 ms; this is a period of micro-break that cannot be tolerated by sensitive electronic equipment.

Series online UPS (double conversion and power conditioned) have the advantage of modern, high frequency electronic processing technology as well as utilising effective isolating power conditioning.

Here, the dc is taken from the battery to drive the inverter, which in turn produces the three-phase waveform required by the connected critical loads. Even with a loss of power from the supply company, the battery remains online and the inverter continues to function.

Uninterrupted power flows to the critical load until the power being restored, or an alternate power source (which could be a diesel generator), can be accessed. This type of UPS combines effective transient protection with the output stability of the online topology to provide the best possible electrical environment for sensitive electronic systems.

The disadvantage is that they are more expensive than the ‘battery in a box’ UPS.

Making the choice

A building’s designer, installer and facilities manager should draw up a list of criteria in order to decide whether to use a dual feeder or UPS backup scheme. Considerations are:

• duration of autonomy of the UPS;
• voltage regulation – ±2·5%;
• frequency regulation – ±1 Hz;
• harmonic distortion level – 5% permissible and tolerable;
• overall energy efficiency;
• site conditions and floor loading requirements;
• the nature and ratio of loads (linear, non-linear);
• redundancy (availability) requirements;
• power quality requirements under lightning and transient switching conditions;
• effects on quality and reliability of installation and commissioning;
• maintainability and connectivity;
• noise levels of the adopted solution;
• emc levels of the adopted solution.

One way to evaluate the various redundant UPS configurations is to look at the mean-time-between-failure (MTBF) or downtime. Using CIBSE Guide K and IEEE failure rates of the critical power load equipment and components (such as cables, transfer switches etc), it is possible to estimate the relative MTBFs of various redundant UPS configurations.

However, duty-holders need to be aware that there will be some reduced reliability caused by the control circuits and human factors in a particular situation. However, the lesson to be learned from this exercise is that the addition of any parallel UPS increases the availability of dependable power, thereby raising the reliability.

Tony Sung, Leon Markwell, Graham Kenyon, Geoffrey Stokes, Paul Harris and Lisa Collins of the CIBSE Electrical Services Group co-authored this article.