The Australian National Electricity Market (NEM) rules and arrangements were developed in the 1990s when the majority of the energy was generated from large thermal generating units that were relatively close to the major load centres. These generation and load centres are also interconnected to allow energy to be transferred and traded to improve reliability and efficiency. New generators have open access to the network provided they negotiate their connection with the network operator and meet acceptable performance standards.
In the past decade changes in technology and climate change incentives have driven an evolution in the NEM rules and market arrangements. However, in recent years pace of change has increased and there has been a dramatic increase in the number of solar and wind generating systems installed and actively being assessed. These generating systems are smaller than the existing thermal generating units and are generally connected in more remote low voltage locations due to resource availability.
The paper will discuss how the NEM has evolved under the high volume of concurrent connection applications, including:
The German energy transition results in an ongoing increase of decentralized renewable energy, especially PV, on the one hand. On the other hand, also an increasing penetration of new electrical loads like power-to-heat systems or electric vehicles can be observed. These factors pose a significant impact with a focus on distribution grids and eventually lead to local congestions.
In consequence, new tools and coordination processes are needed to avoid congestions in advance. Smart markets that can exploit available flexibility in the grid and allocate these for the use by grid operators in a market based manner for congestion management is one possible solution that is examined by FfE within the project C/sells. In order to tap the relevant flexibility potential that has an impact on the congestion, it is essential to get access to available small decentralized controllable systems that are located in the affected grid parts. However, specific prediction of such systems, especially electrical loads like heat pumps, electrical storage heating or electric vehicles is a relevant challenge.
This paper tackles this issue by using historical data for each type of system and calculating the simultaneity factor for a certain analyzed number of small systems over one year. The simultaneity factor describes how many systems are working at the same time. To calculate the simultaneity factors more than 1,500 profiles over one year were generated and analyzed. This process leads to simultaneity factors and the responding confidence intervals for different pool sizes, different times of the day and also temperatures in case of power-to-heat systems. Once this step was done, the available power can be calculated by multiplying the simultaneity factor, the number of systems needed and the average power. Finally, a look-up table was generated to serve the aggregation function of the smart market platform in order to generate pooled flexibility offers.
Furthermore different methods to create appropriate pool sizes were developed. In the paper the different methods were described and the advantages and disadvantages for each method are shown. The size of the pool and the resulting effectiveness finally lead to a conflict of interests as in general the larger the pool, the higher the simultaneity factor, but the more inconsistent the effectiveness factors of the pool.
Finally, the availability prediction is conducted under consideration of external predictions and integration of several factors such as the number of systems, the equivalent daily mean temperature for the case of heat pumps and the type of day. Ultimately standardized aggregated flexibility offers are generated in time steps of 15 minutes that can be placed on the smart market.
The increased variability and uncertainty being induced in the power system with growing penetration of renewable resources necessitates the need for power ow stability studies, analysis and contingency response, to address various operational concerns. The increased variability and uncertainty being induced in power system with growing penetration of RE necessitates the need for power ow studies, stability studies, analysis and contingency response, to address various operational concerns. The present study devises a real time angular instability detection tool based on an adaptive critical angle threshold for voltage angle. The angular separation of the station bus under supervision can be monitored with respect to the critical angle, thus providing a key indicator for initiation of remedial action schemes. The novel scheme is derived from local synchrophasor measurements and doesn’t require knowledge of the entire intricate power system parameters, thus keeping up reliability and computational ease as the key facets. The devised methodology is tested using New England 39-bus system using ETAP/MATLAB.
This paper presents a method for estimating the available active power reserve of a curtailed photovoltaic (PV) power plant composed of multiple inverters. The technique is based on a cyclic algorithm implemented in a plant controller, which calculates and sends active power setpoints to the individual inverters such that some of them operate in Maximum Power Point Tracking (MPPT) mode, while the rest are curtailed to follow a given setpoint at the plant level. When an inverter operates in MPPT mode, its active power measurement is stored in the controller memory, to be used as an estimation of its maximum available power (Pavail). The sum of the individual inverter Pavail estimations provides an estimation of the plant maximum available power. The inverters take turns being curtailed and operating in MPPT mode, allowing the Pavail estimations to be regularly updated such that the global maximum available power estimation accurately follows the variations caused by passing clouds or the sun’s movement. The proposed technique is compatible with existing PV inverter technologies, and can be used to adapt existing PV power plants, enabling them to provide grid frequency control or other ancillary services based on active power flexibility and real-time estimation of maximum available power.
A new setup with an East-West facing horizontal single axis PV tracker array is being tested in full scale by DTU and European Energy, and compared to fixed tilt 25° South facing rows of similar PV panels. The PV panels at the tracker setup automatically follow the sun over the day. Compared to the fixed setup, the PV panels at the tracker setup have higher production in the morning and in the afternoon, where the PV panels are oriented towards East or West respectively, but less production during midday, where the PV panels are oriented horizontally, and the suns incident angle onto the panels are rather high compared to the fixed tilt solution. The total energy productions over the day and over the year are not very different for the two setups. However, the production profile for the tracker setup correlates better with the profile of the typical power consumption, and therefore has a higher value for the power system.
The paper presents comparisons of the values of the production from comparable PV installations at the two different installation setups, based on measurements from the large-scale test installation at DTU. The values of the productions are indicated and compared based on the Nordpool Power Spot Market prices on an hourly basis, valid for Denmark East (DK2).
The number of photovoltaic (PV) installations in the low voltage (LV) grid has increased globally in the last decade. Simultaneously, the sales of electric vehicles (EVs) have rapidly increased. In this paper, an investigation of the impacts of single phase charging (slow charging) of EVs along with single phase PV power generation on a residential electricity grid is provided. This paper also considers unbalanced loads in the electricity grid. Several scenarios were modeled, and each scenario represented a degree of severity of the unbalance in the grid. The results show that the tested grid could maintain the voltages within 0.905–1.103 per-unit (pu) when unbalanced and with PV generation and EV charging loads. The active power losses, however, increased by 42% as compared with the case without EVs. The unbalance of the customer loads results in a 0.03 pu decrease in the minimum voltage in the grid. In addition, the active power losses were higher by 12% compared with the balanced customer load case. In summary, future papers are encouraged to simulate the loads in their unbalanced form, and to try to balance the loads among the phases of the grid.
To address challenges, which come along with the further progress of the “transition of the energy system”, so-called smart-market platforms are part of current research projects (e.g. C/sells as one of the SINTEG showcases). The “Altdorfer Flexmarkt” is the FfE approach to implement a smart-market platform. The goal of ALF is to solve coming challenges at the distribution network. Challenges arise due to additional components and their operating mode at the generation and consumption sector. The technical flexibility of these components (so-called flexibility options) should become accessible and applicable regarding the demand of flexibility from the local distribution network.
The lack of standardized access and the high individual costs constrain the integration of small-scale flexibility into approaches like smart-market platforms. Therefore is to verify, if the infrastructure of smart metering systems (iMSys) is suitable to fix these challenges.
The intention to rollout iMSys in Germany bases on a decision from the EU, which forces the implementation of smart meter in any member state. Germany has decided to implement mandatory smart meter at various groups. This infrastructure is capable to transmit measurement values and control signals, which are necessary for the participation and the internal process regarding the smart-market platform.
To combine smart-market platform and the infrastructure of iMSys a functional requirement analysis is needed. This study compare the requirement of the smart-market platform and the capability of the intelligent metering system. Beside the single components several parties are involved to implement a smart-market platform considering the infrastructure of iMSys.
Demonstration of the concept requires a field test. Since connection user become actively involved in the energy system as owners of potential flexibility options, the demonstration should contain around 50 participants with various types of flexibility. Goal of the field test is the technical proof-of-concept: This concerns the smart-market platform as well as the infrastructure of iMSys. The analysis contains different use cases regarding various types of flexibility options and their capability of external control.
The field test not only comprises technical analyses, but also includes the development of a concept for the integration and accompaniment of involved connection user. For examination, the potential of existing concepts and technologies for the participation of end users and new roles in the intelligent energy system as well as the acceptance of a digitalization of the energy system is part of the concept development.
In summary, the paper describes the opportunity to implement a smart-market platform with the infrastructure of intelligent metering systems. Additional, technical und functional constraints regarding this combination are shown. Beside the technical use cases, the role of user participation is part of the paper.
With the increase in renewable generation sources in the global generation mix, large scale photovoltaic power plants are becoming increasingly ubiquitous. These types of plants are heavily reliant on power electronics in order to mitigate the stochastic nature of renewable energy sources. The increased controllability of the system provided by the power converters, also increases complexity of the control architecture requiring more in-depth knowledge of the systems dynamics. For this purpose, control theory can be used to analyse possible interactions between the different control loops and electrical components.
In this paper a detailed interaction analysis between a PV power plant and power plant controller(PPC) is performed. This study is approached obtaining the small-signal model of the PV power plant, including a specific model of the inverter control, the medium voltage cable layout, a power plant controller (PPC) and the AC grid. This model includes the dynamics associated to the PPC as well as the detailed control of the PV inverters and communication delays. Based on this model, the oscillation modes of the system are highlighted and a participation factor analysis is performed to identify which variables can be problematic. This analysis is carried out for different communication delay times and different inverter operation points. The derived linear model and the subsequent mathematical analysis results are validated through time domain simulations.
Household PV markets are expanding in a rapid pace in developing nations due to the large technology cost reductions achieved during the last decade as well as supporting governmental policies. However, the regulatory context of each country and the emergence of business models inside each market also play a significant role in this diffusion. This paper compares how Indian and Brazilian business models for residential PV market are being structured, either in the grid-connected or in the stand-alone PV markets. The upfront capital investment to purchase and install a PV system is still an obstacle for an average household in the two nations, and this affects business models. We compare features such as type of service, value proposition, revenue model, communication channels, costs and technologies, and discuss how the emerging business models are contributing to enhance access to electricity. Finally, we identify the main barriers for innovative models in each context, including third-party ownership, utility-owned and crowd-funding. We discuss how the current regulatory frameworks support or prevent such models to emerge in both household PV markets.
This research examines, from the lens of photovoltaic (PV) hosting capacity, how centralised control provided by a distribution management system compares with autonomous control. The analysis here includes two study feeders, both with a nominal voltage of 12 kV. Feeder 420 is a high hosting capacity feeder, with a peak load of 5.5 MW, and Feeder 683 is a low hosting capacity feeder, with the same peak load. For each feeder, there are three different types of control equipment: LTC/regulator taps, capacitors, and PV reactive power. The autonomous control utilises the existing local control settings for the distribution assets, and either a constant power factor or a volt-var curve for the PV reactive power. The centralised control has visibility of the whole feeder and chooses the optimal settings for the distribution assets and PV reactive power to maximise the hosting capacity for additional PV on the feeder. An optimisation program, intending to maximise the size of the installed PV unit, calculates the hosting capacity and evaluates feeder constraints using OpenDSS. OpenDSS determines the control settings for assets not being controlled by the DMS. The goal is to maximise PV size without voltage violations or thermal overloads.
Renewable energy forecasts are essential for energy markets and grid stability at different forecast horizons. Machine Learning (ML) models are becoming more popular for forecasting solar energy and are complementing traditional physical models. In this analysis, we are comparing day-ahead solar power forecasts of four different ML methods and two physical models. The dataset includes power measurements of 17 photovoltaic (PV) power plants in Germany and Numerical Weather Prediction model data from the European Centre for Medium-Range Weather Forecasts at a temporal resolution of 15 minutes from April 2018 until June 2019. Standard and optimized system parameters are used as input to the physical model. Within our contribution, we start with a brief description of the different ML methods, the physical models and its parameters, and the dataset. Afterwards, we will present the day-ahead forecast errors of each individual model and compare the best physical model set-up to the best ML model.
As part of the digitalization of the energy system transformation, measuring stations in Germany will in future be equipped with smart meters. The installation obligation applies preferentially to large consumers as well as renewable energy systems (RE) and controllable systems. The Smart Meter Gateway (SMGW) functions as a communication unit in the Smart Meter and provides a universal platform for energy system transformation. SMGW integrate modern measuring equipment across all sectors and record the actual energy consumption in accordance with calibration law. The so-called CLS interface (Controlable Local System) can also be used to integrate and control controllable systems communicatively. In future, the smart meter will thus fulfil two main tasks for the integration of renewable energy systems by recording the feed-in process and implementing the technical requirements for retrieving the actual feed-in as well as for remote control. Even if there is currently no obligation to do so under § 9 EEG 2017 (7), it makes sense to implement this via the uniform infrastructure of the smart meter.
The present article describes the added value of smart meters in the integration of solar, because in order to meet the requirements for the integration of solar systems between 30 and 100 kilowatts in particular, the SMGW initially offers a tariff application case (TAF) for the recording of feed-ins with a resolution of 15 minutes, which are transmitted cyclically.
In addition, this article illustrates how market participants can use the smart meter to map their business models in parallel to network control. The aim is to use flexibility mechanisms to increase the integration of renewable energy systems or storage facilities. The controllable systems provide a flexible share of the performance on the market. In the event of bottlenecks in the affected grid area, a grid operator can request this flexibility according to market rules. In the event that the bottleneck cannot be remedied, grid operators continue to have emergency measures at their disposal, such as feed-in management.
Since high data streams are to be expected in the course of digitization, the processing of which is associated with considerable additional expenditure - e.g. in the form of switching operations -, smart meters can be considered combined with grid automation systems to automate the entire process, thus reducing complexity for humans. As a blueprint for practical implementation, a possible communication and control infrastructure will be presented in the context of this article and evaluated on the basis of field tests gained by the BUW together with its partners SPIE, STEAG, VOLTARIS, the VSE Group and the German Research Center for Artificial Intelligence (DFKI) in the context of the DESIGNETZ research project. In addition, the grid state estimation will present AI methods for PV power prediction that lead to a minimization of switching operations.
Increasing shares of variable renewables in the generation portfolio are resulting in a rapid growth of power electronic converters on the grid, which has implications for the inertial and primary frequency response of power systems. This can be more challenging for (synchronously) isolated power systems, as they typically involve fewer synchronous generation units online, and consequently frequency support services from conventional units can be scarce at certain times. Therefore, it is vital that non-synchronous resources, e.g. wind generation, solar photovoltaic (PV) plant and battery energy storage systems (BESS), contribute towards improving the frequency response through providing short-term (i.e. operating before the primary response timeframe) frequency response control actions. Such measures can enable containment of the frequency nadir within specified limits after contingency events, by managing the high rate of change of frequency (ROCOF) through fast injection of active power.
This paper investigates the use of utility-scale solar PV and BESS, against a large background of wind generation capacity, as fast reserve resources in the All-Island power system (AIS) of Ireland for the year 2030, which has recently committed to a 70% renewables target by this date. The 70% target implies that the system could be operating at, or near, 100% renewables share for extended periods of the year. The Ireland power system operates under a number of technical constraints, and, in particular, the system non-synchronous penetration (SNSP) limit places an upper boundary on the on-line contribution from non-synchronous generation for system stability reasons . The presented work focuses upon understanding the sensitivity of the required speed of response and volume of reserve relative to system conditions. The simulation study is intended to demonstrate the impact of different features, e.g. droop control settings, available capacity, ramp up rate, of short-term frequency response capability from solar PV and BESS on the dynamic performance of the All-Island power system. The investigation provides insights on parameterisation of frequency regulation controls implemented on solar PV and BESS to achieve improved performance.
 J. O'Sullivan, A. Rogers, D. Flynn, P. Smith, A. Mullane, M. O'Malley: “Studying the maximum instantaneous non-synchronous generation in an island system - frequency stability challenges in Ireland”, IEEE Trans. Power Systems, Vol. 29(6), 2014, pp. 2943- 2951.
In a future power system with high penetration of photovoltaic power generation (PV), the power flow in transmission networks would be different from current status. In order to fully utilize the power output of PV installed in suburban area and rural area, the reinforcement of distribution network facilities or the installation of battery energy storage system (BESS) in each distribution network is required. In order to estimate total requirement of reinforcement in all distribution networks in a power system service area, we are conducting to develop a model to estimate electricity demand profile in arbitrary distribution network territory for a year at one hour temporal resolution based on various statistical data sets such as grid-square statistics of National Census. This study using the proposed model assesses the impact of high penetration PV on future residual load profile in consideration of change in population and increase in electricity share in energy demand. Then, focusing on a reverse power from each distribution network in consideration of current substation capacity, this study estimates a required capacity of BESS in order to fully utilize power output of PV.
With 50 trolleybuses and nearly 100 km of catenary, Solingen is the city with the largest operating trolleybus system in Germany. All its operating trolleybuses are equipped with auxiliary combustion engines. However, the public transport system is served with an additional 50 conventional diesel buses.
In project "BOB-Solingen" - the acronym BOB denotes the German words “Batterie-Oberleitungs-Bus” –the entire public transport sector is intended to be completely electrified by integrating novel battery-trolleybuses, which combines proven trolleybus technology with the latest battery technology. The BOB creates the next generation of trolleybuses, where they are able to drive on routes with no catenary as well by means of the included battery.
Moreover, a complete smart trolleybus system is developed at the end of the project, with charging stations for electric vehicles (CS-EV), decentralized renewable power generators such as photovoltaic (PV) systems as well as stationary power storages which are directly connected to the catenary. The stationary storages will consist of obsolete trolleybus batteries to increase their cost efficiency by establishing a second-life utilization concept. The Institute of Power System Engineering at the University of Wuppertal will develop and implement the essential automation system for the DC grid to use its existing catenary infrastructure as effective as possible within its physical limitations.
In order to realize an intelligent operation of the grid, a load flow of the future grid including the above mentioned actuators has to be modelled and simulated. By means of the simulation, critical grid situations can be detected. These might occur more frequently due to the fact that e.g. the additionally implemented actuators can cause an increased number of peak loads as well as surplus power which have to be handled.
This paper intends to achieve a more reliable and efficient overall system, by effectively positioning the stationary power storages. The storage units’ positions will be determined depending on the load demand and the power availability in the system which are heavily influenced by the novel battery-trolleybuses and the PV systems.
Various simulated future scenarios will show the impact of the new actuators to the dynamic grid. In addition, a proposed location finder for the stationary power storage systems will be implemented in order to compare and evaluate several positioning techniques.
The presented work in this publication is based on research activities, supported by the Federal Ministry of Transport and Digital Infrastructure, the described topics are included in the project “BOB Solingen”.
The Hornsdale Power Reserve (HPR) project, consists of a 100 MW Battery Energy Storage System (BESS) in South Australia and was delivered in a period of under 100 days. The HPR project has proven to be a commercial success and a critical part of the Australian National Electricity Market (NEM) providing energy, contingency and regulation services and a critical part of that regions power system protection scheme. Many renewable generators have since identified opportunities to implement BESS projects for additional revenue streams and capture the energy that would be foregone due to system curtailments. However, many of these new potential projects have either failed to materialize, been reduced in scope or have been delayed.
If the Hornsdale BESS project was so successful in such a short delivery time, why have subsequent projects proven to be either too great a financial risk or have been difficult to implement?
Some of the existing problems had been already identified during the HPR implementation but a number of other factors have since emerged that have significantly increased the complexity of these new BESS projects and resulted in either the project abandonment, restrictions or costly delays. Many of the factors that are restricting the potential implementation of BESS projects are not technical. All of the existing battery projects that have been successfully delivered since the HPR project have been from large organizations that have experience with the complexities of power systems and functional trading and operations teams.
BESS installations are a critical missing part of the power system to allow for large proportions of renewable energy in Australia and for many other international markets, but what is required to make these projects a success?
The rapid pace of development of power technology is much faster than the reform processes of market rules and operations. There has been a huge increase in the number of individual generator stakeholders in the market, especially with new renewable generation projects. Market reform is becoming increasingly difficult to achieve with an uneasy balance between incumbent financial interests and fair and efficient market design that matches the reality of the dynamic power system.
It is imperative that the renewable industry can successfully implement BESS projects and that these projects are successful. By mitigating the technical and non-technical risks and by the adoption of reliable and proven systems, it is possible for small and large renewable generators to develop new revenue sources in energy and ancillary service markets by delivering successful BESS projects and participate in the realization of the goal of 100% renewable generation power systems.
Recent development and cost down of PV technology drive change of power system structure. The participation of PV generation is increasing and the size of each PV farm is getting larger. In order to integrate large PV farms into the main grid, substation for interconnection needs to be sized properly. Unlike substations for load and conventional generators, PV farm substation has an uneven utilization ratio due to characteristics of solar radiation. With proper sizing method for the capacity of the substation can reduce the building cost of facilities. A combination of an energy storage system can further reduce the capacity of the substation. Battery Energy Storage System(BESS) can shift the peak production of PV during the daytime to midnight. According to market circumstances, BESS can reduce further construction costs by producing profit based on time difference of electric cost. For proper sizing of substation capacity, several factors must be considered including environmental factors, market structure and BESS in the system. In this paper, a series of assessment methodology is introduced to calculate the optimized capacity of substation and BESS for PV farm interconnection. The long term solar radiation data is analyzed for a given site of the PV farm. Based on market structure, the operation of BESS is optimized to make maximum profit during operation. The iterative calculation of each step results in the calculation of the optimized capacity of BESS and substation for given PV farm size.
Photovoltaic (PV) technology is considered as a very promising solution for electricity generation in the quest for combating climate change. However, several grid issues are emerging when its share in the electricity generation mix increases significantly. A promising way to combat this, is the installation of battery energy storage systems (BESS). This paper presents the project Erigeneia which aims to develop a new advanced energy management system (EMS) algorithm in order to support the increase the PV hosting capacity of the grid. The EMS will incorporate the functions of energy management, frequency and voltage regulation and power smoothing at both residential (decentralised) and community (centralised) levels. Furthermore, a battery sizing algorithm and forecasting tool will be developed. The solutions will be implemented and tested at five residential storage systems in Cyprus and at one central storage system in Turkey.
Dynamic RMS model validation against commissioning test measurements of any newly installed or upgraded generating system is required in the National Electricity Market (NEM) of Australia. Many variable renewable energy plants, particularly PV solar, have been commissioned recently in Australia and much experience has been gained through the model validation process.
Typical compliance tests include assessments of the power plant voltage control (fixed reactive power / voltage / voltage-droop / power factor), active power dispatch control and frequency control. The tests used for demonstrating compliance of the solar farm (or wind park) against the technical performance requirements set out in the Generator Performance Standards (GPS) can be used for model validation purpose.
This paper describes the process of validating the dynamic RMS models by comparing the simulated responses against actual plant responses and AEMO’s model accuracy requirements. A few challenges from the model validation process, including impacts of reduced number of inverters, frequency disturbance simulation method, solar irradiance and low PPC meter sampling rate, are discussed. Some example model overlay plots are presented for illustration.
Harmonic emission limitation is a power quality compliance requirement under the Australian National Electricity Rules (NER). Many solar farms have been commissioned recently in Australia and much experience has been gained during commissioning tests. Unlike a conventional synchronous machine, solar PV power plants use inverter-based technologies which generate a wide range of harmonic frequencies. Total harmonic voltage distortion measured at the point of connection (PoC) depends not only on the harmonic current profiles of the inverter, but also the harmonic impedance of the external grid and the solar farm collector network impedance. the latter can be calculated from the Norton equivalent impedance of the inverter. Unfortunately, the information required for harmonic calculations is often difficult to obtain. This paper presents a harmonic filter design methodology, which has demonstrated during commissioning to reduce total harmonic distortion and individual harmonic voltages to allocated levels. Additional topics including IEC summation law, impacts of cloud shading and filter switching, and statistical measurement requirements are also discussed and illustrated with field measurements.
Power optimiser are becoming increasingly popular in residential photovoltaic installations. Besides monitoring and safety advantages the power optimise can increase production under partial shading conditions due to the individual maximum power point tracking performed independently for each panel. This study quantifies the production gains with power optimisers through both empirical and theoretical city-wide simulations. Empirical measurements on adjacent string and optimiser PV systems showed that shading losses from a tree decreased from 17% to 13 %. A study of 1100 single family houses in an urban suburb showed that 2/3 of the houses would have a gain with optimisers above 20 kWh/kWp and year.
In the EU-founded project OSMOSE, RTE and Ingeteam will install a demonstrator (Ringolab) to show the technical and economical viability of building a grid-forming function upon a traditional energy storage system. Although main theoretical results have already been validated on real hardware in a laboratory environment within the framework of the MIGRATE project, a grid-connected demonstrator represents a step further toward the standardization of grid-forming converters. In this paper we present its technical description and we show stable association of the grid-forming control with different
DC side power sharing and energy management strategies while considering a hybrid energy storage system.
This paper aims to propose a methodology allowing the assessment of the flicker contribution of a photovoltaic (PV) installation. Due to the increasing presence of PV in the electric power grid, challenges arise to ensure power quality. One of the challenges covers the increase of flicker levels that can be caused by the fast power fluctuations of PV. Grid operators become suspicious about the possible exceedance of flicker limits but, in contrast with wind turbine parks, no standard method exists to evaluate this for PV installations. In this paper, a methodology is proposed that can be used for this evaluation. The methodology incorporates a decision tree which allows deciding between different approaches based on the available data and some key variables.
The rapid expansion of photovoltaic (PV) systems has raised overvoltage concerns. This paper investigates voltage variations measured for four hundred on-line PV installations in Sweden. The specific production (SP) is dependent on the PV array sizing ratio as inverters will curtail energy in an oversized PV array and prevent high SP values. Small (<10 kW inverter size) three-phase residential solar PV systems showed the least impact whereas single-phase systems showed the most impact for the same amount of power injected per phase. Solar PV systems were grouped based on the post code location into urban and rural areas. The urban areas were found to be more resilient to voltage rise as the solar PV increases as voltage in urban areas fluctuates within a narrower voltage band in comparison to the rural areas. The spring season in which the highest voltages occurred is more challenging to solar PV integration. The winter season has the lowest solar PV energy production and has negligible effect on the voltage rise, yet SP values up to 0.8 p.u can be reached.The rapid expansion of photovoltaic (PV) systems has raised overvoltage concerns. This paper investigates voltage variations measured for four hundred on-line PV installations in Sweden. The specific production (SP) is dependent on the PV array sizing ratio as inverters will curtail energy in an oversized PV array and prevent high SP values. Small (<10 kW inverter size) three-phase residential solar PV systems showed the least impact whereas single-phase systems showed the most impact for the same amount of power injected per phase. Solar PV systems were grouped based on the post code location into urban and rural areas. The urban areas were found to be more resilient to voltage rise as the solar PV increases as voltage in urban areas fluctuates within a narrower voltage band in comparison to the rural areas. The spring season in which the highest voltages occurred is more challenging to solar PV integration. The winter season has the lowest solar PV energy production and has negligible effect on the voltage rise, yet SP values up to 0.8 p.u can be reached.
With large shares of variable renewable electricity (VRE) generation in the energy system, there is an inevitable demand for power balancing. This study compares three system setups. Two setups contain a local combined heat and power (CHP) unit with district heating (DH), the first also have central thermal energy storage (TES) together with heat pumps (HP) while the second have central electrical energy storage (EES) and no HPs. These systems are denoted DH/TES and DH/EES, respectively. The third setup denoted EES/HP does not include DH, instead HP and domestic hot water (DHW) heaters are used at end user together with central EES. The degree of electricity balancing (reductions of electricity surplus and deficit), CO2-emissions, and economic factors, have been calculated for the three setups.
The heat and electricity demands for a small community (111 single family homes) in Sweden were simulated. The national power supply was assumed to be 100% renewable with large share of VRE (60% wind power and 10% solar). The power balancing demand profile was estimated to illustrate when there were times of electricity surpluses and deficits to which the local community could contribute with balancing power. The residential area had building integrated photovoltaics, PV, installed with 150kWp.
In the DH/TES, the thermal production unit was a small-scale CHP unit together with HP and a TES. The CHP was operated to meet the electricity load demand. If there was an electricity surplus, the HPs used this to produce heat. All produced heat was supplied to the DH network if there was heat demand present. Any surplus heat was stored in the TES. At times without a power balancing demand (neither surplus nor deficits) the heat demand was supplied from the TES.
In the DH/EES, the thermal production was a small-scale CHP unit (no HP). The CHP was operated to follow the heat demand with co-generation when the electric and thermal demand coincided. During electricity surplus, the EES charged, and during times of deficits it discharged to supply unmet electricity load. Since the thermal demand at all times was supplied by the CHP unit’s production, co-generated electricity from CHP where given priority over usage of the EES.
In the EES/HP, the heating demand was supplied with air-to-air HP and DHW electric heaters at end users. Neither DH nor local CHP was applied to this setup. The EES was the same as for the second setup.
The results show that the DH/EES reduced the electricity balancing demand the most, but had the highest CO2 eq-emissions. The EES/HP had no emissions, but the highest cost. This setup also increased the balancing demand. The DH/TES performed most satisfying when electricity balancing capacity, CO2 eq-emissions, and costs were weighed together.
Renewable energy resources have been widely introduced to power grids. However grid-connected inverters for RES have no rotating machine characteristics, including inertia and synchronizing power. Power system stability may decrease at the time of power system disturbances. Virtual synchronous generator control (VSG) methods applied for the inverter have been suggested as a countermeasure for power system stability. The authors proposed in their previous paper that a VSG control scheme based on the synchronous generator swing equation is applied for active power control and an automatic grid voltage control (AVR) scheme is applied for reactive power control, and those VSG and AVR controllers are used together for the grid-connected inverter. PSCAD simulations show that the suggested method is effective for improving stability. This paper describes the results of experiments for validating the proposed VSG control scheme through hardware in the loop simulation (HILS) composed of a small-capacity inverter and a real-time simulator OPAL-RT.
The amount of solar PV installed capacity has steadily increased to 44.5 GW at the end of FY2017, since the introduction of the Feed in Tariff (FiT) to Japan in 2012. This seems to be more likely to achieve 64 GW, the target value of solar PV generation in the 2030 energy mix by the Japanese government. On the other hand, since the first curtailment of solar PV was conducted on October 13th, 2018 in the Kyushu area, the curtailment has been frequently executed including wind power after that. In order to reduce curtailment, it may be possible by changing the operation. For example, it shall review about a priority dispatch rule in each area, a demand response, and so on. However, it will be difficult to execute immediately, as it relates to other power generation and consumer circumstances as well. Therefore, in this study, it will be focused on cross-regional interconnector and pumped hydro energy storage (PHES).
In Japan, there are 9 electric power areas which connected each other by cross-regional interconnectors. According to the historical operation, cross-regional interconnectors were secured as emergency flexible measures, but after the implicit auction was started from October 2018, it is used on merit order. There are many time slots when market split occurred between electric power areas because the capacity of the cross-regional interconnector is not enough. An enhancement of cross-regional interconnectors is required. Regarding PHES in Japan, they have been built with nuclear power plants for several decades. Because the output of nuclear power generation is constant, so PHES is used to absorb the surplus at nighttime when the demand declines. All nuclear power plants in Japan have been shut down after the accident at the Fukushima Daiichi Nuclear Power Plant following the Great East Japan Earthquake that occurred on March 11th, 2011. There are several nuclear power plants that have been restarted (9 reactors, as of August 2019).
In this study, the amount of curtailment for solar PV in the Kyushu area is sent to the Chugoku area using the cross-regional interconnector (Kanmon line). Then, PHES in the Chugoku area is pumping with low price. Because the spot price in the market is low when the curtailment is executed. After that, PHES is generating at night with high price when the solar PV is not generating. It makes a profit by the deference for the cost of pumping and the revenue of generating by PHES. As a calculation result, for one week from May 2nd to 8th, 2019, a profit becomes 152.2 million JPY (about 1.22 EUR). For this purpose, it is necessary to raise the operation capacity of the cross-regional interconnector up to the rated capacity with the frequency control function of solar PV instead of the capacity to keep frequency in the event of an accident. This will allow the further introduction of solar PV in Japan.
Abstract – This paper introduces a cooperative decentralized control strategy to control power management in a cluster form of DC microgrids integrated with multiple energy Renewable energy sources (RES), Energy storage systems (ESS) and distributed generations (DGs). Control strategy is based on a hierarchical structure of primary, secondary and tertiary controllers. DC bus voltage signal is used to enable power sharing among different power sources and for seamless transition between modes of operation. Primary control is mainly based on an adaptive droop mechanism to regulate voltage output of each converter and balance generation amongst sources. Secondary controller is applied to measure voltage across the micro grid and update voltage set point for primary controllers. Consequently, Tertiary controller is a higher level controller, that is responsible for the economic dispatch of all power sources in the whole cluster. A cooperative control strategy is applied to implement the primary, secondary and tertiary controllers. Recent research papers, had implemented tertiary controller in a single mode dc microgrids to control voltage setpoints and perform economic dispatch of power generation. However, novelty of this paper is in implementing tertiary controller in a higher level to deal with a cluster form of dc microgrids to reduce the entire generation cost by controlling voltage output of every single power converter. Ultimately, the effectiveness of the proposed technique is verified using MATLAB/SIMULINK 2016b based on a composite dc microgrid test system. Keywords – DC microgrid, Decentralized control, Tertiary controller, Renewable Energy Sources (RES), Energy storage systems (ESS), Distributed Generation (DG), Economic dispatch ….
Facing large and further increasing amounts of solar photovoltaic (PV) capacity, transmission and distribution system operators (TSO and DSO) need to include high quality and regionally optimized forecasts of upcoming PV power feed-in into their planning processes. The energy and cost efficient integration of many of systems into national and international electricity grids requires reliable knowledge about the expected feed-in with lead times of hours, days and weeks ahead.
Aiming to generate a benefit to the combination with other established state-of-the-art forecasts, our model has been developed primarily for German TSOs, but is applicable to any region where PV power plants are known to feed in. Based on TSOs’ valuable feedback we will present an overview of our alternative model approaches and the results we achieve. Not only the optimal weighting of best numerical weather predictions (NWP) leads to notable benefit in operational forecast ensembles, but the unique modelling of emerging features like self-consumption and storage of produced PV power.
Our model architecture optimally accounts for local weather phenomena as well as the relevant parameters of distributed PV power plants. We will compare the improvements of mixing NWP and refining the forecast model steps on the way to generate valuable grid node forecast and to reasonably differentiate between potential and actual PV power feed-in forecasts.
This contribution will describe the current and emerging challenges of forecast users as well as our experiences in identification of most promising solutions. We show the use-case of a German TSO and present the improvements in day-ahead forecasting for the year 2018.
At higher solar penetration levels, grid integration becomes challenging due to inherent solar production forecast uncertainty. Minimising the risk and cost associated with reserve generation, demand response and energy storage requires not only accurate forecasts, but sharp and reliable probability information to inform conservative decision making under uncertainty. Between 15 minutes and 4 hours ahead, probabilistic cloud “nowcasts” can be made using the current observed cloud state from satellites, resulting in a narrowing envelope of solar radiation uncertainty at shorter lead times, which can inform utility decision making and help drive automated systems. Traditionally, this sort of technology has been developed on a case-by-case basis for a particular region or utility, on a multi-year project basis, typically at a high cost to the utility and/or taxpayers, and with high barriers to entry for prospective new or additional users.
This paper and presentation will demonstrate a global, probabilistic, rapid-update nowcasting system for solar radiation data and PV power which can immediately provide live (real-time) and nowcast data globally through a web API. This solution allows programmatic, automated data requests for easy integration into testing and operational systems. The API framework, coupled with an instant-access “Freemium” business model, is designed to encourage adoption by Engineering and Data Science teams solving real problems associated with solar radiation and solar power production.
This submission will describe the Solcast approach to generating semi-dynamical probabilistic forecast outputs, via ensemble cloud advection that uses modern machine learning methods and background weather state information. Ensemble performance is monitored in real-time to produce a weighted forecast mean, as well as ensemble spread at key probability levels for decision making.
Finally, we will present two real-world case studies from our user base, which provide useful examples of the power system applications of this live nowcast service. The first case is improved production forecasting for a fleet of utility-scale solar plants, the second case is improved electricity demand forecasts accounting for the impacts of collective behind-the-meter PV production.
Full decarbonisation of energy sectors in Australia can be achieved through large-scale adoption of renewable energy in the electricity sector, along with direct or indirect electrifications of heating, transport, manufacturing and mining. A synergy of large-scale interconnection between electricity grids, energy storage, and demand-side management enabled by smart energy system can help to build energy system resilience with high level of energy security and reliability. Remarkably, large-scale integration of solar energy with support from wind enables 100% renewable energy to be cost-competitive with fossil and nuclear energy.
Japan has been experiencing rapid PV deployment since feed-in-tariff program was launched in July, 2012 one year after the Great East Japan Earthquake and Tsunami. While, the PV penetration is affecting in each of the balancing area in Japan according to the level of the penetration, (Fig.1)
Since October 2018, due to the penetration, frequent PV curtailment have been occurring in Kyushu area, in one of the four main Islands in Japan. The reason of the curtailment is not the congestion of transmission lines, but the demand and supply balance of the total area. The situation in Kyushu of the high shares of PV and non-carbon generation with limited interconnection capacity is comparable with that of Ireland with high penetration of wind. The power system operation with high share of non-dispatchable generation (PV+wind+geothermal +nuclear) is enabled through sophisticated operation planning and real-time operation including conventional and variable-speed pumped hydro plants.
Although the Kyushu have been keeping the security of supply even with the marginal system operation, there were several cases in which they experienced extremely large PV forecast error. The experiences showed the necessity of new flexibility reserve of large volume in order to maintain the sustainable variable renewable energy deployment.
Under the situation, the paper discusses the latest status of PV curtailment in Japan and emerging requirement for new operational reserve against extreme PV forecast errors and provision the flexibility including utilization of new distributed resources such as EVs and heat-pump water heaters. The new requirement and supply of new flexibility is analyzed by demand and supply balance including the newly-developed general flexibility model including contribution of PV and wind output control and demand responses.
In this paper, starting from a review on grid code requirements for different sizes of PV-systems, typical challenges of utility PV- and storage plants, like low SCR (Short Circuit Ratio) operation and examples for the correct use of existing functions like dynamic voltage support in case of grid faults are depicted. Not all challenges arising with a further increasing share of large PV-systems in the coming years may be handled with the existing control functions. New control schemes like “grid-forming” inverter controls are widely discussed to overcome such challenges in weak or low inertia grids with a high penetration of power electronic generation and loads. In this paper, motivations for the application of voltage control mode or “grid-forming” control schemes are discussed. Such control schemes are applied today in PV-diesel-hybrid-systems with a very high share of PV, in order to maximize the use of energy from PV by allowing to turn off the diesel gensets. Following the same principle physics, battery storage systems are expected to become an important building block in the future’s interconnected electrical energy system. Some examples of the performance of grid-forming inverters connected to the transmission or distribution-grid are thus presented. Depending on the application, connecting the storage system to the DC- or the AC-side of PV-Systems may be favorable. In other cases, a stand-alone storage system connected to the grid may provide the best solution. A general insight on such topological approaches is given in the paper. Finally, there is an outlook on open e.g. regulatory or grid-code related challenges to be resolved.
In recent years, the penetration of renewable energy systems such as photovoltaics (PV) has progressed rapidly in Japan’s distribution system. The government provides assistance to install PV by purchasing surplus generated electricity and implementing a feed-in tariff (FIT) system since 2009. Under the FIT scheme, if a renewable energy producer requests an electric utility to sign a contract to purchase electricity at a fixed price and for a long-term period guaranteed by the government, the electric utility is obligated to accept this request. Purchase at a fixed price for amounts less than 10 kW is guaranteed for 10 years from the start of the program. The consumer expired this program will appear since 2019.
Japan’s distribution system includes voltage management equipment such as load ratio control transformers (LRT) and step voltage regulators (SVR). Because their outputs are weather-dependent, voltage control assumes more importance and proves to be a complex problem. To this end, the collaborative control method is considered for the distribution system equipment and power conditioning sub-system (PCS). The PCS suppresses voltage rise when the output voltage deviates from the upper limit. This function consists of normal operation, reactive power output, and output suppression. In a distribution system including a PCS equipped with this function, deviation from the upper limit of the distribution line voltage can be avoided, but the generated power may be suppressed.
One solution to this problem is to extend the waiting time limit of the voltage rise suppression function by considering coordination with an SVR . Typically, the waiting time limit of the SVR is set to 45 sec at a minimum. By setting the operating time limit of the PCS to longer than 45 sec, the SVR can first optimize the voltage and reduce the amount of output suppression using the PCS voltage rise suppression function. However, voltage control performance is degraded by extending the operating time of the PCS. Thus, a method without a waiting time limit is considered to be effective in reactive power control, wherein the power generation opportunity loss does not occur in the voltage rise suppression function.
Hence, in this paper, we present the voltage rise suppression function of a PCS that can cooperate with an SVR and also cope well with short cycle output fluctuations.