What causes Losses?
Raw losses values and losses rates have to be calculated over long periods (at least 3 years) to ensure stability and robustness, as a losses for a given year may not be significant due the variability and uncertainty such as data collection hazards, climatic conditions. In order to see the effect of reducing losses WPD need to be able to determine the baseline level of current losses.
Distribution network losses can be broadly defined as the difference between the electrical energy entering the distribution network, from base generation or embedded generators either upstream / same level / downstream networks, and the electrical energy exiting the distribution network, for consumption purposes and properly accounted for it, in percentage terms for a particular period.
Distribution network losses are typically broken down into three categories: -
- Technical losses;
- Non-technical losses;
- Other factors.
Technical losses fall into two areas fixed losses and variable losses: -
The total amount of technical loss is made up of a fixed component (a function of the network itself, independent of the load on the network) and a variable component which is dependent on the level of load on the network. Variable losses may also be impacted by the power factor, network imbalance and the effects of harmonics.
Some electrical energy is dissipated by network components and equipment such as transformers or conductors as a result of being connected to the network and being energised. Even if no power is delivered to customers, the system has losses just because it is electrically energised. These losses take the form of heat and noise and are called ‘fixed losses’ or ‘no-load losses’, because they are independent of how much electrical energy the network delivers.
Two types of losses in transformers are known to exist: -
- ‘Iron losses’ are losses that stem from the reversal in magnetic polarity of the steel in transformer cores in every AC cycle. This causes the material to pulse (which emits a humming noise) and to heat up.
- ‘Copper losses’ are losses that stem from the circulation of induced currents in conducting parts that are not copper windings, such as the iron body or steel core of the transformer.
Besides transformer inefficiency, another source of fixed losses is the electrical insulation in network equipment. Imperfections in electrical insulation lead to the flow of very small currents across them in transformers, overhead lines, underground cables, and other network equipment. These types of fixed losses are called ‘dielectric losses’ or ‘leakage current losses’. Corona losses, a particular case of these type of losses, they occur in high voltage and extra high-voltage overhead lines. They vary with the voltage level, the physical wire diameter, and with weather conditions such as rain and fog.
While fixed losses do not change with current, they depend on the applied voltage. However, as the applied voltage is relatively stable while the network equipment is energised, they are essentially fixed. Therefore, fixed losses are a function of the network itself and depend mainly on the number of energised components. In general, fixed losses contribute too roughly between a quarter and a third of the total technical losses on distribution networks. There are a number of smaller effects which can also contribute to technical losses on the network:
The variable component of losses is created by the heating effect of electricity passing through the cables and windings. All conductors, whether they are coils in transformers, aluminium or copper wires in overhead lines or underground cables and even in switchgear, fuses, or metering equipment, have an internal electrical resistance which causes them to heat up when carrying electric current. As a result, the variable losses change as power flows increase and decrease (proportionally to the square of the current), transmission networks experience a lower level of losses because at higher voltages a lower current is required to transmit the same amount of electric power. Additional factors such as the effect of network imbalance, power factor and power quality can also have an impact on variable losses, as they influence the value of the currents flowing through the conductors.
Additionally, variable losses are also dependent on the length and the cross section of the network line as they vary in proportion to the conductor resistance. The resistance of a conductor decreases as its cross sectional area increases. Therefore, the effect of losses is reduced with larger cable sizes. A similar principle also applies to the variable losses in transformers, where the cross sectional area of windings, and the materials used in them, influence the variable losses.
In general, variable losses contribute roughly between two-thirds and three-quarters of the total power system technical losses. They either aim to lower the system power flows or to lower the resistance of the transportation paths. A reduction in the utilization levels of network assets can contribute to lower both current and resistance. Any higher capital investments required for loss reduction must show a positive lifetime cost benefit analysis.
Non-technical losses are caused by actions that are external to the power system. They refer to lost energy that is not directly related to the transportation of electricity and occur independently of the physical, technical characteristics of the network (technical losses). Cases of non-technical loss cannot be fixed by upgrading equipment or altering network design. Instead investigations, audits and collaborations with other bodies are required. This kind of loss involves the abstraction of electricity with a loss of revenue to both the network operator and the supplier.
There are two main types of non-technical loss: -
Theft in Conveyance
There are several ways in which electricity can be taken from the network illegally. Theft and fraud generally account for a majority of the non-technical losses from the network. These are important challenges for the DNO, and require a concerted effort from a range of stakeholders to mitigate them. It is difficult to gauge the exact extent of this type of losses as a large proportion of it is likely to go undetected.
When illegal connections to the network are made; properties do not have a meter installed or a registered supplier, it is referred to as theft in conveyance.
Not all supplies in distribution networks are metered. There are many items of electrical equipment where it is neither practical, nor cost-effective, to measure energy consumption using conventional meters. In these circumstances, there are legitimate unmetered supplies whose energy demand is estimated rather than accurately metered. All unmetered connections can be treated as any other type of load, provided that it is registered, properly estimated and accounted for. Moreover, customer-related unmetered connections (e.g. public lighting) or some of the DNO’s own consumption (e.g. auxiliary services of substations) can be adequately contracted from an energy supplier and paid for by regular tariffs as any other normal consumption. Therefore, unmetered consumption, whether related to customers or the DNO, can be excluded from non-technical or technical losses, respectively, provided they are adequately contracted. Only the difference between the real and estimated unmetered consumptions is part of non-technical losses.
In the case of equipment such as street lighting, traffic lights and road signs it is not practical to meter every unit. Instead bills are estimated using the power rating of the equipment, the approximated time of use and the number of units. It is not uncommon for these estimates to be inaccurate or an inventory of equipment to be out of date. In order to reduce these losses, DNO’s must work alongside customers with unmetered supplies to improve the accuracy of inventories, to produce more accurate bills.
Other factors that affect network losses are: -
- Phase Imbalance;
- Power Factor.
A network which does not have its load evenly distributed across all three phases will have higher currents in at least one phase meaning it is not optimised for losses. There will also be currents flowing in the neutral conductors if they are present. Due to the quadratic dependence of losses on current, this load imbalance across the three phases will increase losses.
Imbalance is found on all parts of the low-voltage (LV) network due to customers who use one or two phases having different load consumptions. On the high-voltage (HV) network, imbalance is due to the uneven distribution of single-phase transformers or two wire spurs and different loads on each phase for three-phase customers. The most obvious way to reduce phase imbalance is to carefully balance the aggregated load on each phase, but as customer consumption is not always predictable and varies at different times of day, this can be difficult.
A rural LV overhead network could be rebalanced across phases relatively simply by moving the overhead service connection to a different phase of the overhead main. This is more difficult on an urban underground LV network, as this requires existing service joints to be excavated and new joints made to move customer supplies to different phases.
Interventions to alter connections will help balance customers and load across a network based on the maximum demands of those customers. Balancing load profiles over time is very complex, so some imbalance will always occur at certain times of the day. Loads will change in the future so any action taken to balance the network will have to consider what changes are likely to occur in the future.
Harmonic effects are essentially distortions to an AC current profile. They can occur in transformer windings because the AC magnetising current is not perfectly sinusoidal. However, this usually occurs on the triple harmonics (3rd, 6th, 9th etc.) so on a normal three-phase system they are all in phase and do not result in any real harmonic voltages. However, if other equipment connected to the network produces harmonics they will not cancel in the neutral conductor. These can then cause additional I²R losses, as in real terms the losses formula becomes I²R+√H where H=harmonics on the network, this increases the overall load on the network which in turn increases the losses.
There are two ways to define the power in a system. The real power is the capacity of the system to do work. The reactive power is the product of the voltage and the current flowing. The power factor is the ratio of the real power to the reactive power. Where the power factor is less than unity the current has to increase to deliver the required amount of real power, which results in a loss. This has historically been an issue for installations used by industrial and commercial customers, where most motor loads or power electronic loads were seen. Developments in domestic power electronics and heat pumps mean networks will start to see this issue occurring more on the LV Mains networks.
Since 2010 WPD have been including an excessive reactive power charge for HV and LV half hourly metered, via the Use of System Charges, where customers have a power factor of 0.95 lagging, this is to ensure that the reactive power is kept to the minimum as with any load the DNO has to cater for the reactive power for the sizing of the circuit even though that reactive power is not being used effectively.