Is it time to genuinely adopt 230 V as our distribution voltage?

The article in September/October 2008 Electrical Solutions raised many questions regarding supply voltage variances. Author Paul Stathis was warned that he could be opening a can of worms. The advice he received was not wrong. In this article, I will attempt to answer some of the issues raised in the previous article and bring attention to more issues that need to be addressed by the electrical industry.

The problem faced in Australia is that, while electricity distribution network service providers (DNSPs) run their networks with a nominal system-voltage of 240/415 V, increasingly the equipment connected to it is designed for 230/400 V operation. Although this may not seem a major difference, it is in reality a ‘silent killer’, causing the premature failure of much electrical equipment and costing the community dearly.

History

Edition 6 of the standard that defines voltages, IEC60038, introduced 230/400 V in 1983, in an effort to introduce a standard international low-voltage value. The rationale was to assist manufacturers to only have to produce one range of products - a major step forward in improving the overall economics of products.

Prior to this, AS2926 applied the same voltage range in Australia as the UK: 240/415 V ±6%. When the 230/400 V compromise was reached, it was decided to widen the voltage tolerance such that all existing tolerances would be included. Accordingly, the standard became 230/400 V ±10%.

IEC60038 Ed 6 stated that existing 220/380 V and 240/415 V systems shall evolve towards 230/400 V by 2003. During this transition, 240/415 V systems should bring the voltage within the range +10%, -6%, after which the tolerance of 230/400 V ±10% should have been achieved. It was then assumed that the voltage range would be tightened back to ± 6%.

New IEC60038

The recently published IEC60038 7th edition notes that the transition to 230/400 V systems has been completed in many countries. However, 220/380 V and 240/415 V systems still exist. The note regarding tightening the tolerances has been deleted altogether, replaced it seems with an ‘informative’ Annex defining the highest and lowest voltages at the supply and utilisation terminals for a 230/400 V system with a tolerance of ±10%.

The IEC Committee responsible for this standard (TC8) doesn’t consider voltage tolerances its responsibility but has just provided ‘informative’ data. The utilisation voltage is on the basis of a maximum allowable voltage drop of 4% within an installation in accordance with the IEC equivalent of our Wiring Rules (IEC60364-5-52:2001). Consideration is presently being given to allow the voltage drop to increase to 5% for power circuits in line with our Wiring Rules.

Tolerances

This raises the question of how voltage tolerances are set. There are no internationally agreed rules that define ‘allowed’ steady-state voltage range. Similarly, there are no agreed Australian rules that prescribe the allowable steady-state voltage range. How could this be so?

Despite the fact that the Australian electricity market is mostly regulated under the National Electricity Market framework, the regulation of technical performance standards for DNSPs is under state jurisdiction.

The National Electricity Rules are defined loosely, stating that provided the power factor of the load at the connection point is within agreed limits, voltage at the ‘connection point’ should remain within ±10% of the ‘normal voltage’.

Accordingly, despite the existence of AS60038, a DNSP can elect to declare its nominal voltage as being 240 V and state that the voltage range is 240 V±10% (216-264 V). Some states regulate the low voltage, while others allow DNSPs to make their own declaration of voltage.

Curiously, our power quality standards don’t deal with the fundamental aspect of steady-state voltage. Although these standards deal with many power quality issues, such as harmonics, flicker and unbalance, apart from AS60038, which simply sets voltage levels, there is no standard yet developed that addresses the issues of steady-state voltage. What is the acceptable range? How is the voltage measured? What is the measurement period? The EL34 Australian Standards Committee (of which I am a member) is presently working on developing a new standard to address this gap. I have also been pressing for this to be addressed at IEC level.

Compatibility

There’s a fundamental compatibility discrepancy between voltages that can appear on the network and voltages that equipment standards prescribe to demonstrate their adequate performance. Utilities want as wide a voltage range as possible in order to remain ‘in tolerance’. Whereas, equipment manufacturers want the voltage range to be as tight as possible in order to optimise designs and not have to over-design equipment to cope with extreme voltage ranges. This is contrary to the compatibility curve that is fundamental to the IEC61000 standards.

The major problem is that for steady-state voltage, the two bell curves are reversed. The system voltage deviation from nominal can be +10% or even +14%, while the equipment standards only prescribe testing of the equipment at ±6% or less.

In practice, it doesn’t quite work out that way, as reputable manufacturers want to ensure their equipment lasts in service at least to the end of the warranty period. So manufacturers tend to subject their equipment to more onerous tests than those required by standards. However, consumers have no way of knowing whether the equipment is designed to cope with more onerous conditions than those prescribed by standards. Some form of Star Rating for appliances based on power quality and energy efficiency would be a good step forward.

Equipment performance

So if equipment still performs when large voltage variations occur, why should we be concerned? The problem is that most equipment is designed to operate optimally at a particular voltage. When voltage deviates from the ‘nominal’ voltage, performance is affected, depending on technology used. The simplest example is incandescent lamps (principle also applies to CFLs that are replacing them, due to capacitance, but more on that later). Light output and life expectancy of incandescent lamps are critically dependent on voltage. For 240 V globes, if the voltage is low, light output drops dramatically. If it’s high, say 254 V, output increases dramatically but lamp life typically halves.

Capacitors are affected similarly, where their service life is reduced by the ratio of nominal voltage divided by the overvoltage to the power of 9. Capacitors are used extensively in modern electronic equipment: switchmode power supplies, CFL lamps, VSDs, etc. If manufacturers select minimum-cost components rather than those rated at levels suitable for the highest likely voltage, equipment life will be greatly affected by excess voltage.

So CFLs, which are mostly designed for 230 V, will fail prematurely when subjected to the overvoltage that may be experienced on Australian electricity networks.

Motors are also designed for nominal voltages. If subjected to high voltages, the magnetic flux in the steel circuit will saturate and cause heating of the steel laminations. Under low-voltage conditions, motors will need to draw higher currents to achieve the required torque.

The problem is that equipment standards are based on testing equipment at their nominal voltages. In some standards, equipment is tested for a range of voltages, but rarely is that range greater than ±6%. In any case, the focus of most standards is to ensure product suitability in terms of electrical safety and fire risk. Actual performance, except for energy efficiency, hasn’t been the focus of product standards for many years. Adequate performance is left to the ‘market’ to sort out.

Only ‘prescribed (or declared) items’ must be subjected to testing. Manufacturers can voluntarily have their products approved it they’re not ‘prescribed’, but this tends to be the exception rather than the rule. Most electronic products aren’t prescribed items and therefore don’t undergo any approval process. So Australia is essentially an open market for importers to sell any products they wish with little restraint, apart from the requirements of the Trade Practices Act. In Europe, EMC directives relate to emissions and immunity. In Australia, the only regulatory requirement relates to high-frequency emissions.

Disturbingly, Australian manufacturers and importers aren’t too concerned. After all, shortened equipment life means an early end-of-life, so long as failure doesn’t occur within the warranty period. Given our ‘disposable produce’ mentality, the poor consumer simply buys a new one.

So what’s the problem? Worldwide, low-voltage products are being designed to operate on 230 V systems. According to IEC60038 Ed 7, the maximum system voltage is 253 V. In Australia, although system-voltage is stated to be 230 V, our network is still a ‘240 V’ system. The nominal output voltage of distribution transformers when rated voltage is applied to the high-voltage side is 250/433 V. Under light load conditions, such as in off-peak hours, the biggest problem faced by customers is high voltage on the low-voltage network. Until such time that there is a move by manufacturers, consumers or organisations representing the community, there are no drivers for DNSPs to consider changing and reducing the normal steady-state voltage by 10 V.

After all, there’s a cost impost on DNSPs to reduce the output voltage. If the net community benefit of reducing voltage is positive, regulators need to allow DNSPs to accommodate the cost involved in reducing system voltage. In the meantime, the community needs to accept that any appliances designed for 230 V will continue to fail prematurely and mysteriously.