Bart Verhelst

Introducing small storage capacity at residential PV installations to prevent overvoltages

Low voltage distribution feeders are designed for unidirectional energy supply from transformer to consumer. However, the implementation of small-scale PV production units on local utilities may result in bidirectional energy flows. The simultaneous power injection at sunny moments may cause a serious voltage rise along the feeder. These overvoltages may not only damage critical loads but also switches PV inverters off causing loss of green energy at the most productive moments. This paper presents a method to limit the voltage rise by introducing small battery buffers at local production sites. A smart inverter decides whether the PV energy is injected in the grid or buffered in the batteries. The relation between battery buffer size and overvoltage reduction is presented for a typical Belgian residential distribution feeder. The influence of the buffer along the feeder is calculated by working with synthetic load profiles and solar irradiation data.

Preventing overvoltages in PV grids by integration of small storage capacity

The rapidly growing amount of distributed generation in low voltage distribution grids issues some undesired side effects. Simultaneous power injection may cause a serious voltage rise along the feeder. These overvoltages may not only damage critical loads but also switch off PV inverters causing loss of green energy at the most productive moments. This paper discusses the origin and different possibilities to limit this induced voltage rise. The option of introducing local energy buffering is further elaborated. The relation between buffer size and overvoltage reduction is presented for a typical Belgian residential distribution feeder. The influence of buffers along the feeder is calculated by working with synthetic load profiles and solar irradiation data.

Economic evaluation of the influence of overvoltages and the integration of small storage capacity in residential PV-installations

Implementation of decentralized production units (PV-installations) on low voltage distribution feeders may result in bidirectional energy flow. Simultaneous power injection may cause voltage rise on the feeder. Due to regulations inverters need to switch off when reaching a certain voltage causing loss of production. Because this happens mostly on bright, sunny days production loss can be significant. This paper presents an economic comparison between a situation without production loss, with production loss due to overvoltages and the introduction of storage. The production is calculated based on synthetic load profiles and solar irradiation data. The economic feasibility is calculated based on realistic prices and taking into consideration all available government funding.

Measurement-based modelling and validation of PV systems

This paper presents the analysis and results of modelling of various photovoltaic (PV) systems. Two general models are discussed and presented: an analytical model and an equivalent circuit model, both formulated for main PV technologies currently available on the market. Analytical model does not require any PV system specific input data or parameter, and is formulated as a generic performance model of a considered PV technology. Equivalent circuit model, however, requires specific input data and adjustment of the model parameters, in order to provide an accurate representation of a modelled PV system. The paper provides direct comparison of models based on manufacturers specification data and available measurements, as well as the discussion of obtained results.

Electrical balancing potential in Belgian residential installations

The growing number of prosumers cause an upwards injection of electrical current on the public distribution grid. When the number of volatile electrical production increases, several problems on the electrical distribution system appear e.g. increasing line voltage, unbalance, reverse power flow. So far the necessary amount of electrical power is determined by electrical companies and supply is adapted by regulating large production facilities. Balancing supply and demand on a local level can avoid the injection of electrical current in the public distribution system. This is called Demand Side Management (DSM). Thereby a prosumer can avoid certain power quality problems such as high line voltage, which can lead to a shutdown of volatile electricity production. This paper focuses on the potential for local DSM in a residential building.

Prediction of yield of solar modules as a function of technological and climatic parameters

A solar module is exposed to numerous factors which influence the total yield. Besides the yield being dependent on both the installed technology and placement conditions, some parameters have a direct effect, others have an indirect influence on the actual output. In this study the yield is predicted as a function of both technological and climatic factors. To evaluate the impact of a single parameter, data of a photovoltaic test field with four different technologies and three different test set ups is analyzed to distract an equation for the instantaneous yield of a string of solar modules. An expression is derived to determine the yield based on the ambient temperature instead of the much used panel temperature. This study does not calculate the total yield of the PV system nor does it take into account invertor efficiencies. The yield is calculated at DC side.

Influence of grid configuration on current conducting behaviour in PV installations

On the roof of an industrial site a 385 kWp PhotoVoltaic installation is operational. When production of this system reaches 60% of the installed power, the circuit breaker trips. At sufficient production, measurements show a high distortion of phase voltage and variable waveform of both phase voltage and current. Analysis of the installation showed that a Yy0 transformer is used introducing a high zero sequence impedance. Unbalance in the injected current combined with a high zero sequence impedance leads to a high neutral-ground voltage and distorted phase-neutral voltages. In this paper it will be shown that the tripping of the circuit breaker is caused by the measurement method of the device. This paper analyses the practical measurement results, causes of errors and the solution to the stated problem.

Influence of grid-connected inverters on the power quality of the distribution grid

Many inverters are quoted as having a PF of better than 0.95 when in reality, the true power factor is between 0.5 and 0.75. This paper describes the influence of grid-connected inverters on the distribution grid. The impact of the voltage distortion from the inverter to the distribution grid is relative small without grid impedance. When the grid impedance is involved, the distortion will become higher and so, more harmonics will be injected into the grid. The impact of the current distortion is especially high at low power output, but decreases fast at moderate or high power. If the background distortion of the grid increases, there will be an increase of the voltage and current harmonics. The specific limits of the EN 50160 and IEC 61000-3-12 could be exceeded at such high pollution.

Tripping of circuit breakers in PV installations due to zero sequence field impedance

On the roof of a wool processing company a 385 kWp PV installation is installed. When production of the installation is more than 60% of the rated power the circuit breaker trips. Measurements show high THD(U) of phase voltages and variable waveform of both voltage and current. Analysis of the installation showed that a Yy0 transformer is installed introducing a high zero sequence field impedance. Unbalance in the injected current combined with a zero sequence impedance leads to high voltage drops and distorted phase voltages. Measurements also show high RMS values of the current in the neutral conductor. Analyses show that the tripping of the circuit breaker is caused by the measurement method of the device itself. Higher current values are calculated due to multiple zero crossings of both voltage and current. This paper gives the practical measurement results, causes of errors and the solution to the stated problems.