Measurement of Power Quality Problems Caused by Common New Loads
ABSTRACT
In recent years, two significant factors have greatly influenced energy usage in Finland: the greenhouse effect and rising energy prices. These factors have led to the introduction of numerous energy-efficient devices in the market, such as heat pumps and energy-saving lamps. Additionally, reductions in manufacturing costs have made various new types of electrical loads available. These emerging devices have altered the patterns of electricity usage and the demands placed on the distribution network. Originally designed to accommodate different patterns of electricity consumption, the distribution network may now face increased Power Quality Issues as these new devices become more prevalent.
This paper explores Power Quality Problems that arise from the use of ground source heat pumps and a wood splitter. The analysis is based on measurements conducted within an actual low-voltage distribution network in Finland.
The study reveals that power quality issues can occur when using these modern electrical loads, with a particularly significant issue being the high starting current of the wood splitter. This high current causes a voltage drop, potentially leading to flicker.
I. INTRODUCTION
Energy consumption patterns have shifted in recent years due to the introduction of new types of loads into the distribution network. Three primary factors have driven these changes: the greenhouse effect, rising energy prices, and reduced manufacturing costs of devices.
Efforts to mitigate the greenhouse effect and global warming primarily focus on reducing carbon dioxide emissions. Consequently, the market has seen an influx of devices designed to lower emissions, such as compact fluorescent lamps. Replacing incandescent bulbs with these energy-efficient alternatives is one approach to reducing carbon emissions, as compact fluorescent lamps produce light more efficiently. European Commission Regulation No. 244/2009 mandates the gradual phase-out of incandescent bulbs from the market.
The rising cost of energy has prompted consumers to invest in devices that reduce both energy consumption and costs. A common example of this behavior is the replacement of oil heating systems with heat pumps, or the addition of an air-to-air heat pump to complement electric heating. In Finland, government support has accelerated this transition. Figure 1 illustrates the rapid increase in heat pump installations in Finland from 1996 to 2008. This trend is mirrored across Europe, with heat pumps becoming the predominant heating system in single-family homes in Sweden, where they account for approximately 34% of all heating types.
Figure 1. Number of installations of different heat pump types in Finland during the years 1996-2008.
Different types of heat pumps impact electrical energy consumption and usage patterns in various ways. For instance, replacing an oil heating system with a ground source heat pump increases a household's overall electrical energy consumption. Conversely, adding an air-to-air heat pump to an existing electric heating system can decrease consumption. However, in both scenarios, when comparing the previous primary energy consumption to the new electrical energy usage, the overall energy consumption decreases since the majority of heating energy comes from the ground or air. This reduction in primary energy consumption leads to a decrease in carbon dioxide emissions, contingent upon how the electricity is generated. The greater the reliance on fossil fuels for electricity production, the more significant the emissions reductions achieved by using heat pumps.
Heat pumps exemplify loads that have gained popularity due to technological advancements and reduced manufacturing costs. In Finland, an air-to-air heat pump, including installation, now costs between €1,200 and €3,500, making it an affordable option. Ground source heat pumps, which are more expensive due to the need for a ground circuit and higher power capacity, are another example of a new load gaining popularity due to lower costs. Similarly, wood splitters, used for splitting thick logs for fireplaces or sauna stoves, have become increasingly common due to their affordability and convenience.
These new loads are altering how electrical energy is consumed. Despite their potential to reduce overall energy consumption, these devices can also introduce power quality problems due to their distinct electrical characteristics. The design principles of distribution networks, which are becoming outdated, do not always meet the requirements of these new loads. As these loads become more widespread, power quality issues are likely to become more frequent.
This paper examines the power quality problems associated with the use of ground source heat pumps and wood splitters. The devices were selected based on power quality complaints received by a distribution utility. The study is grounded in practical case measurements conducted in real distribution networks in Finland. The paper first provides a theoretical overview of power quality issues related to the aforementioned loads, followed by a detailed description of the case study measurements. Finally, the results and conclusions of these measurements are presented.
II. THEORETICAL BACKGROUND OF POWER QUALITY
Power quality is defined as "a set of parameters defining the properties of power quality as delivered to the user under normal operating conditions, in terms of continuity of supply and voltage characteristics (symmetry, frequency, magnitude, waveform)" [7]. This paper focuses on power quality in terms of voltage quality. The limits for voltage quality are defined in the standard EN 50160, which outlines the voltage characteristics of electricity supplied by public distribution networks. The standard aims to define and describe supply voltage characteristics concerning frequency, magnitude, waveform, and line voltage symmetry. These characteristics are subject to variations during normal system operation due to load changes, disturbances from specific equipment, and faults caused mainly by external events. Such variations are random in both time and location, meaning that, on rare occasions, the limits may be exceeded .
The EN 50160 standard defines voltage characteristics for low and medium voltage networks [8]. This paper only considers voltage characteristics in low voltage networks, focusing on flicker caused by rapid voltage changes, voltage levels, and voltage dips, as these are common causes of power quality complaints. Figure 2 shows the distribution of power quality complaints received by a distribution utility in Finland between 2003 and 2005. Approximately 70% of all complaints were related to voltage changes.
Figure 2. Distribution of power quality complaints in one distribution utility in Finland during the years 2003-2005 [9].
Flicker and voltage level disturbances are often caused by devices that require power to operate. When a device is connected to the grid, it draws current, causing a voltage change over the distribution network's impedance. This phenomenon is described by Equation 1. Depending on the device's connection type, it may affect phase voltages differently. For example, if the device is connected in a three-phase configuration and is symmetrical, all phase voltages will experience similar effects, either a voltage drop or rise. However, if the device is connected in a single-phase configuration, the phase voltages may experience different voltage changes due to star point displacement; for example, one phase voltage could rise while the others drop.
ΔU = vector of voltage change ZN = vector impedance of the distribution network IDEV = vector device current
The EN 50160 standard sets permitted levels for flicker, stating that 95% of long-term flicker severity in any week should be less than or equal to 1. For voltage levels, the standard specifies that 95% of the 10-minute mean r.m.s. values of the supply voltage should be within the range of Un ± 10% during any period, and all 10-minute mean r.m.s. values should be within the range of -15% < Un < +10%. Exceptions may occur in remote areas with long feeder lines or networks not connected to a large interconnected system, where voltage levels could fall outside this range. In such cases, customers should be informed of the conditions [8].
III. THEORETICAL BACKGROUND OF HEAT PUMPS AND WOOD SPLITTERS
Both heat pumps and wood splitters rely on induction motors for power, making these motors the most significant component in terms of current usage and power quality. Wood splitters typically use a single-phase induction motor, while heat pumps may use either a single-phase or three-phase induction motor, depending on the type. Larger heat pumps, such as ground source heat pumps and air-to-water heat pumps, generally use three-phase motors, while smaller heat pumps typically use single-phase motors.
The most significant impact of an induction motor on the distribution network occurs during startup when the motor draws a high starting current. This high current causes a voltage change across the network's impedance, as illustrated in Figure 3 and Equation 2. Figure 3a presents the equivalent circuit of a polyphase induction motor, while Figure 3b presents an approximate equivalent circuit. The approximate circuit assumes that the reactive component of impedances z1 and zm is much greater than the resistive component, and that the voltage E2 is only slightly smaller and nearly in phase with voltage V. These assumptions hold true for conventional induction motors in the normal running range. Equation 2, derived from Figure 3b, shows that when the slip of the induction motor is small, the motor draws a high current from the network. The slip is 1 at the moment of starting the motor and decreases to near 0 after startup. Single-phase induction motors have a different equivalent circuit and equation for current draw compared to polyphase motors, but the high starting current effect is similar in both cases [10].
Figure 3. a) Equivalent circuit of a three-phase induction motor. b) Approximate equivalent circuit of a three-phase induction motor [10].
r1 = resistance of the stator x1 = leakage reactance of the stator xm = magnetizing reactance r2 = resistance of the rotor x2 = leakage reactance of the rotor s = slip of the motor
The starting current poses significant power quality challenges because it reaches its peak value at startup and remains relatively stable during normal operation. From a power quality perspective, the frequency of motor startups is critical. If the startup frequency is high, more voltage changes will occur. The duration of heat pump operation depends on the heating energy demand, the pump's size, and its parameter settings. If the heating demand is close to the pump's nominal capacity, the pump may run continuously for long periods. However, if the demand is significantly lower than the pump's capacity, frequent startups and shutdowns may occur. For instance, one ground source heat pump supplier specifies that the minimum time between shutdown and restart is 10 minutes.
For wood splitters, startup frequency depends on the device's operating logic. There are two primary operating logics for moving the hydraulic piston of a wood splitter. The first involves performing all piston movements with hydraulic control while the induction motor is running. The second involves using hydraulic control for the pressing action while the motor is running and a spring for the backward movement when the motor is stopped. The second logic results in a much higher startup frequency. In our measurements, the wood splitter we studied required a startup each time a new log was split.
IV. DESCRIPTION OF PRACTICAL CASE MEASUREMENT STUDY
Our study involved two practical case measurements conducted in Finland, one in the rural area of Lempäälä and the other in the urban area of Tampere. The scope and environment of each case were slightly different. We used Dranetz PX-5 and Dranetz 4400 power quality analyzers as measuring devices. This section provides a detailed description of both practical case measurements.
Lempäälä Rural Area Network
In Lempäälä, the study focused on power quality problems caused by a wood splitter in a rural area network. The wood splitter was a single-phase device with a nominal motor power of 2.2 kW. The splitter's operating logic required a startup each time a new log was split. The low-voltage network consisted mainly of aerial bundled cables (AMKA) with cross-sections ranging from 70 mm² to 35 mm². The rated power of the 20/0.4 kV transformer supplying the network was 200 kVA. In this network, calculated single-phase short-circuit currents at customer locations ranged from 148 A to 1,400 A. We studied power quality problems caused by the wood splitter at three locations with measured short-circuit currents of 146 A, 275 A, and 350 A at customer supply terminals. There was a 10 m extension cord between the customer supply terminal and the wood splitter's plug point, resulting in short-circuit currents of 136 A, 200 A, and 233 A at the plug point. Figure 4a provides an overall picture of the low-voltage network in Lempäälä, with measuring locations along feeders 1 and 2. Measurements were taken at the plug point, the customer supply terminal, and near the transformer while the wood splitter was operating. Additionally, when the wood splitter was operated at the location with a short-circuit current of 146 A, measurements were also taken at the location with a short-circuit current of 275 A, as both locations were on the same feeder. The measured quantities included voltage and current waveforms and parameters defined in standard EN 50160, except for the short-term flicker severity, which was measured over a 5-minute period.
Figure 4. A) Overall picture of the low-voltage network in Lempäälä. B) Overall picture of the low-voltage network in Tampere with measuring point locations.
Tampere Urban Area Network
In Tampere, the study focused on power quality problems caused by ground source heat pumps in an urban area network. The low-voltage network consisted mainly of underground cables with cross-sections ranging from 120 mm² to 300 mm², and the rated power of the 20/0.4 kV transformer was 315 kVA. In this network, calculated single-phase short-circuit currents ranged from 445 A to 9.7 kA, and three-phase short-circuit currents ranged from 1.1 kA to 10.8 kA, indicating a relatively strong network. At one end of this low-voltage network, two terrace houses replaced their shared oil heating system with separate ground source heat pump systems. Two heat pumps with nominal heating powers of 25 kW and 36 kW and maximum electrical powers of 9.9 kW and 13.2 kW, respectively, were installed in each terrace house. The pumps were started directly online, resulting in high starting currents each time they were activated. This location was selected for the study due to customer complaints about flicker. Figure 4b provides an overview of the low-voltage network, customer supply terminal short-circuit currents, and measuring points. The measured quantities included voltage and current waveforms and parameters defined in standard EN 50160.
V. PRACTICAL CASE STUDY RESULTS
This chapter presents the results of the practical case studies, organized similarly to the previous chapter for ease of reference.
Lempäälä Rural Area Network
The wood splitter caused significant power quality problems at the customer end. Each log split resulted in a high starting current relative to the short-circuit current, causing a significant voltage dip in the phase to which the wood splitter was connected at the customer supply terminal. As a result, the most severe problems were related to flicker and the frequency of voltage dips. Figure 5 illustrates the waveform and RMS value of the starting current during one wood splitting event at a location with a short-circuit current of 275 A at the customer supply terminal.
Figure 5. Waveform and RMS value of one wood splitter startup at a location with a short-circuit current of 275 A at the customer supply terminal.
It was observed that the phase voltages of the other phases, not connected to the wood splitter, increased at the customer supply terminal due to star point displacement in the low-voltage network, as predicted in Chapter III. Figure 6 illustrates this effect, showing the phase voltages at the customer supply terminal during a wood splitter startup. As seen in Figure 6, the phase voltage in the connection phase drops so dramatically that each startup produces a voltage dip according to standard EN 50160 [8].
Figure 6. Phase voltages at a location with a short-circuit current of 275 A at the customer supply terminal during a wood splitter startup.
The overall network impact results of the practical case study in Lempäälä are summarized in Table 2. In measurements 1 and 3, the wood splitter was connected to phase L2, while in measurement 2, it was connected to phase L1. Table 2 shows a significant increase in the short-term flicker severity index at the customer supply terminal in every phase when the wood splitter was in operation. This indicates very noticeable flicker, meaning that even short-term use of the wood splitter would exceed the long-term flicker severity limit of Plt=1, as defined in standard EN 50160 [8]. It should be noted that an electric chainsaw was also used during the wood splitter operation, which slightly increased the short-term flicker severity index. However, the chainsaw's impact on phase voltage was much smaller than that of the wood splitter.
Table 2. Overall results of the practical case study in Lempäälä.
In addition to the results shown in Table 2, it was observed that when the wood splitter was operated at measurement location 1, power quality problems were also recorded at measurement location 2, which was 250 meters away. The operation of the wood splitter increased the short-term flicker index at the other measurement location to 7.2 in the connected phase and to 2.8 and 1.1 in the other phases, even though the wood splitter was operating at measurement location 1. Despite the significant power quality problems observed at the customer end, no issues were detected at the transformer. One way to mitigate these issues is to use only wood splitters where the piston movements are controlled hydraulically while the induction motor runs continuously.
VI. CONCLUSIONS
This paper examined power quality problems caused by increasingly common electrical loads, based on practical case measurements conducted in a real distribution network in Finland. The study focused on two groups of loads: wood splitters and heat pumps.
The wood splitter, a single-phase device with a nominal motor power of 2.2 kW, was tested at different locations within a rural low-voltage network in Finland. Calculated short-circuit currents in this network ranged from 1.4 kA to 146 A, with measured currents of 148 A, 275 A, and 350 A at the customer supply terminals. The wood splitter caused significant flicker problems due to its high starting current and frequent startups of its induction motor. Flicker issues were observed across all three phases due to star point displacement from the operation of the single-phase device. The flicker problems also extended to nearby customers along the same feeder. One way to prevent these issues is to allow only wood splitters where the piston movements are hydraulically controlled while the induction motor runs continuously, reducing the number of startups.
In Tampere's urban area network, power quality problems caused by heat pumps were studied. Four large heat pumps were installed in two terrace houses. The startup of these pumps caused the short-term flicker severity index to exceed the irritation threshold of 1. Additionally, the startup of heat pumps in one terrace house increased the flicker index in the other terrace house, indicating cross-disturbance. Although the short-term flicker index occasionally exceeded 1, the long-term flicker index remained below 1, indicating no flicker problem according to the standard. Nevertheless, the flicker was clearly visible during heat pump startups, particularly in lighting, which could be irritating to some customers. As a result, the heat pump supplier will install soft starters to address the issue.