8th Solar & 17th Wind Integration Workshop // E-Mobility Integration Symposium 2018 (SIW / WIW / E-Mobility 2018 // Stockholm, Sweden)
Stockholm / Sweden, 15 to 19 October 2018

Solar Integration Workshop 2018

Improvements for Online Feed-in estimations based on PV power output measurements
Submission-ID 002
Daniel Lassahn, Niklas Riewald
meteocontrol GmbH, Germany
To estimate the accumulated PV power feed-in of an region, measurements of reference PV systems can be used. Since the data of a perfectly representative ensemble of systems is hard to obtain, statistical post-processing methods can be used to correct resulting estimation errors.

In this paper the \textit{Analog Sampling} algorithm is introduced to correct diurnal error patterns by taking the solar angle into account. It is a local regression method that uses the Analog Ensemble metric for fitting linear models on localized subsets of the data. The algorithm is evaluated on the basis of measurements from meteocontrol GmbH and real feed-in values from the Tennet control area. It can be shown that it is able to correct errors resulting from the use of a predominantly utility scale reference ensemble.

Allocation of Frequency Control Reserve with Micro Grids Participation for Power System Security
Submission-ID 008
Norhafiz Salim 1, 2, Takao TSUJI 2
1 1. Universiti Teknikal Malayisa Melaka (UTeM ), Malaysia
2 2.Yokohama National University, Japan
This proposal investigates the technical features of yielding frequency control reserve (FCR) potential based on multiple micro grids composition. It has been reported by many academic articles and much is already known about its technicalities but only some have investigated their technical impacts and leveraged its benefits.It is a big challenge to aggregate micro grids as such to govern the frequency control. During contingencies, for instance, loss of generating unit or transmission facility, transmission system operator is committed to maintaining proper amount of active power in reserve for compensation purposes. The frequency and tie-line power interchange are very susceptible mainly with load flexibility for an interconnected area. Balance control and delta control are some strategies being used when involving renewable energy resources (RES) and distributed energy resources (DER) participation in controlling the network frequency. However, the frequency control basis for the emergence of micro grids to support on primary frequency control is still unclear on how to efficiently integrate them without introducing technical instabilities and violating operational limitation set by a system operator. In this research, a comprehensive frequency control mechanism that aimed to suppress frequency deviation and tie-line power interchange is presented. In particular, the proposed approach using Flat Frequency Controller (FFC) and Tie-Line Bias Controller (TBC) could offset for any supply and demand imbalance after disturbances; minimizing frequency deviation and preserve inter-area power dispatch. Extensive simulation with different micro grid components, load characteristics, and control parameter will be carried out to outperform a relevant start-of-the-art controller. The effectiveness of proposed controller will further be verified in Malaysian Electric Power System (MEPS) and the results of this research are beneficial to the design consideration in FCR and LFC mechanism for Malaysia power system perspective.

Sensitivities in Hybrid Energy Systems
Submission-ID 009
Mike Alexander Lagler, Ernst Schmautzer, Robert Schürhuber
Graz University of Technology – Institute of Electrical Power Systems, Austria
Motivation and central question

A hybrid energy system intelligently combines energy generation and distribution systems as well as energy storage systems for electricity, heating and cooling. By combining thermal and electrical systems, resources can be spared, energy costs can be saved, and the perceived comfort can be increased.

However, this coupling requires optimal cooperation of the electrical and thermal system components, which are sensitive to internal and external influences (e.g. usage, geographic and topological situation), technical design (e.g. old, renovated, new), economic constraints and ecological priorities (e.g. local or global aspects).

In this work, the sensitivities of various parameters on a hybrid energy system are investigated:

Since consumption and generation of energy do not coincide in every time step, it is without the use of energy storage or load management necessary to purchase or sell energy at suboptimal tariffs, or to adapt the load profile (e.g. consumer behaviour) accordingly.

Methodical approach

This work presents a hybrid simulation model, developed with the software MATLAB, which couples the electrical and thermal subsystem models of a single-family house. Based on this model, the optimal use of the distributed energy generation and storage systems is determined based on defined scenarios by means of linear optimization (MILP - Mixed Integer Linear Programming) taking into account the above mentioned dependencies.

The optimization is performed in a temporal resolution of 15 minutes over a period of at least one year. Sensitivity analyses (e.g. the influence of the weather, size and operation of the electricity/heat/cold storages, load profiles, quantity- and time-dependent tariffs) are performed and evaluated for the economically optimal operation determined from this (initial situation).

Outlook and conclusions

The aim of the presented work is to investigate the effects of various influences on the economically optimal operation of a hybrid energy system (single-family house) and thus to show the dependencies and weights of the hybrid energy system, whereby these are graphically displayed for easier interpretation.

Daily global solar radiation prediction using artificial neural networks: a new approach
Submission-ID 015
Youness El Mghouchi
Moulay Ismail University, Meknes, Morocco., Morocco
In this study, for determining the best-input scenarios of the used input parameters, a new approach for predicting Daily Global Solar Radiation ( DGSR ) based on a new developed Artificial Neural Networks (ANNs) is presented . Recorded data from 35 stations belonging to different climatic zones were reported for the training and testing of the network. The used input parameters include geographical coordinates, sun declination, day length, day number, clearness index (KT), Top Of Atmosphere (TOA), average ambient temperature (Ta), maximum temperature (Tmax), minimum temperature (Tmin), difference temperature ( Δ T), temperature ratio (TR), relative humidity (Rh) and wind speed (Ws). The obtained results revealed 128 best-input scenarios, while the most relevant input parameters are KI, Ta, Δ T, TR and TOA. This result indicates that the best input scenario for predicting DGSR is based only on three climatological parameters: KI, function of Ta f(Ta) and TOA. In addition, based on these found best-input scenarios and on the least square regression (LSR) technique, 128 new linear relationships between DGSR and the found best-input combinations were developed. The statistical analysis expressed through statistical criteria indicated perfect correlations and approximations between the predicted and measured values of DGSR.

Renewables integration grid study for the 2030 Japanese power system
Submission-ID 017
Rena Kuwahata 1, Peter Merk 1, Tatsuya Wakeyama 2, Steffen Rabe 3, Dimitri Pescia 4
1 Elia Grid International, Germany
2 Japan Renewable Energy Institute, Japan
3 GridLab, Germany
4 Agora Energiewende, Germany
Japan boasts a decent share of renewables (RE) in its power mix: 15% of annual electrical energy is supplied by RE including hydro as well as solar and wind. Policy targets penetration of RE to reach 22-24% of power production by 2030. There are also concrete plans for major electricity sector reforms like unbundling and transition to market-based approaches to facilitate the policy goals. However, there is concern about the way the Japanese power grids will handle the efficient integration of RE owing to the practices for grid security assessment and management currently applied.

International experience has shown that: the growth of renewables can happen unexpectedly fast; and several technical solutions (which are not yet common in Japan) for grid management can be safely applied. However, few RE grid integration studies exist in the public domain, that attest these aspects of the Japanese power grid. The main inhibitive reason is the inaccessibility to asset data and information about operational practices of incumbent grid companies, that would be necessary to perform meaningful studies.

With this background, the Japan Renewable Energy Institute (REI) attempts to bring together various Japanese power industry stakeholders to gather data, validate modeling and simulation results as well as apply progressive technical assessment methods for integrating RE. Based on grid modeling and simulation, grid stability and power flow were tested for high RE penetration scenarios foreseeable in 2030. In this way, the impact of RE on interregional power exchanges and frequency stability is assessed.

In this context, the study contributes to the demonstration of the impact of options (which may not be fully considered by utilities and the regulator today) in efficiently integrating large amounts of renewables and ensuring a robust development of the power grid.

Hybridization of Floating Solar PV with conventional Hydro Power: A cost effective energy storage solution 
Submission-ID 018
vijay Madhusudan
Energy and Energy Consultants, India
Hybridization of Floating Solar PV with conventional Hydro Power: A cost effective energy storage solution

It is a well-known fact that energy is one of the most influential factors for economic, social and industrial development of Nations. The latest trend amongst Nations is to invest in renewable energy technologies that are clean and green. This facilitates them a bit hedge against fluctuating cost of energy and moreover reduces per capita emissions. We understand that each renewable energy technologies have its own share of issues; therefore, the objective of this research paper and study is to identify techno-commercial benefits of hybridizing Floating solar PV with conventional Hydro Power Plants.

Large-scale adoption of Solar PV installations that contributes a significant share in the energy mix of the electrical grid system comes with associated issues like grid instability etc. Few such issues being addressed in this paper includes variable production of energy with seasons, mismatch in generation versus consumption patterns and grid instability. These issues can be largely mitigated with an energy storage solution. We have tried to quantify how a certain issues and disadvantage of one technology is mitigated by hybridizing the two technologies. In our research, we have accounted for infrastructure requirements, analyzed generation patterns from Hydro Power Plants and Solar PV Plants, built financial models and calculated Levelised Cost of Electricity (LCOE) for the technologies to conclude with our results. With this research, it can be proved beyond doubt that hybridization of Floating Solar PV Plants with Hydro Power Plants not only mitigates issues associated with the technologies, but also provides a great opportunity to create low cost energy storage solution and grid stability which is need of the hour in Asia-Pacific.

We are working on a pilot project “ Development and installation of solar-hydro pumped storage in remote village” to ensure 24x7 electricity and inclusive growth of villagers under Technology Mission Division of Ministry of Science and Technology, Government of India.

Optimizing Energy Dispatching in Spatio-Temporal or Time Series Weather Forecasting Modelling, Through Artificial Neural Networks
Submission-ID 019
Ignacio Smith
1, Trinidad and Tobago
UNFCCC, Trinidad and Tobago
A lot of research has been conducted and a lot time has gone into developing models to forecast weather patterns. The approach has usually been the use of large data strings through Time Series and Spatio-Temporal data. However, Deep Learning offers a different approach, as well as it the potential to single handlely revolutionized the short term weather forecasting.

This paper centers on the application of deep learning (Convolutional Neural Networks) and in its dfifferent architectures in the recognition of cloud patterns to allow better forecasting and subsequenty, impoved dispatching of energy generated by PV Plants. Our research points to the advantages of using Convolutional Neural Networks and N-I-N (Network In Network). The results exemplify the advantages of using Deep Learning for weather pattern forecasting and the impact it may have in commodity markets..

Keywords: Deep Learning, Convolutional Neural Networks, Time Series.

Feasibility study of grid connected of 20 MW PV power plant for a village in Chlef region, Algeria
Submission-ID 029
Mohamed Dekkiche
Mechnical department, Chlef University, Algeria
This paper presents an economic, technical and environmental aspects for constructing 20 MW installed capacity grid connected photovoltaic power plants in a village at Chlef region, Algeria. The meteorological data (global solar radiation, wind speed, wind direction, air temperature and relative humidity) were measured at Ouled Fares site in Chlef, from 1 January 2015 to 31 December 2017. The energy load requirement for a village of 10000 habitations is carefully estimated to be ten thousand times of a daily load data measured in one house. Available converters and photovoltaic modules are chosen for this study. The data and the components were used to conduct the feasibility analysis using HOMER software in terms of financial parameters for the site, energy production and greenhouse gas (GHG) emissions. The simulation results found that, the optimal grid-connected system meets the village energetic needs and injects into the grid more than the usually purchased electricity. For the case study, the optimal architecture system has 20000 kW of PV modules, 12200 kW of converter and 4000 kW of grid. It produces 49,874,023 kWh/yr with 77.9% of the energy are produced by the photovoltaic generator and 22.1% are purchased from the grid. The system is able to meet energy requirements load of 20,000,000 kWh/yr of the village with 2.09% energy in excess about 1,040,000 kWh/yr. The simulation results show that the optimal PV/connected system has an initial cost of 26.9 $M and a low NPC of 8.21 $M with COE 0.0135 $/kWh. The obtained results lead to the conclusion that grid connected photovoltaic energy systems are very favorable for Chlef region.

Accurate Method for Active Power Reserve Allocation by Utility-Scale PV Power Plants
Submission-ID 031
Vahan Gevorgian
NREL, United States
For utility-scale PV power plants to be able to maintain the desired regulation range or spinning reserve levels, the plant controller must be able to estimate the available aggregate peak power that all the plant’s inverters can produce at any point in time. The available power is normally estimated by an algorithm that considers solar irradiation, PV modules I-V characteristics and temperatures, inverter efficiencies, etc. However, this conventional method has many uncertainties, depends on availability of accurate system models, and does not account other factors such as panel soiling because of dust. The proposed method can determine the available peak power of the PV plant and maintain desired reserves with high levels of accuracy without use of external sensors or devices. The existing plant hardware and controls can perform this task after addition of the new optimized control algorithm in power plant controller software. We demonstrate and validate this new method using data from the real utility-scale 50 MW PV power plant located in Western United States. The results of sstudy how significant improvement in plant maximum power evaluateion achevieng less than 1% accuracy during even highly variable cloudy days.

Power System planning with Battery Storage in large scale solar integration into Sri Lankan Grid
Submission-ID 036
Ubesiri Narayana Sanjaya, Dr. H.M. Wijekoon
Ceylon Electricity Board, Colombo 02, Sri Lanka
Sri Lanka is geographically located close to the equator, so that it receives an abundant supply of solar radiation during the year. At present, 50MW large scale solar plants have been commissioned and it is proposed to connect around 400MW Solar Plants by 2021 including a 100MW large scale solar plant in Monaragala District of Sri Lanka

With high-penetration levels of renewable generating sources being integrated into the Sri Lankan national grid, conventional generators are being replaced and grid inertial response is deteriorating. This technical challenge is more severe with photovoltaic (PV) generation because of the higher intermittency levels which raise concerns in maintaining grid stability. A Battery energy storage system (BESS) provides a substantial support to enhance the initial response and primary frequency regulation of the power system. Therefore, a requirement arises to investigate the above impacts as well as the improvements through BESS in power system planning perspective.

To investigate this, IEEE 39 bus power system was adopted in the software tool PSSE developed by Siemens PTI and a dynamic model of a solar PV Plant was incorporated with the system. Based on the measurements obtained at a 10MW Solar PV Plant in Sri Lanka, a practical irradiation profile on a high PV variability condition was applied to the Solar PV model to identify the frequency dynamics behaviour. Then, a BESS dynamic model was incorporated with the above solar plant to identify the improvements of the system behaviour under the above PV variability condition. Finally, this setup was applied for Sri Lankan power system in order to integrate the 100MW large solar plant in Monaragala District into the Sri Lankan Grid in 2021.This paper discusses the Transmission System study and its outcomes on the dynamic behaviour of power sytem with Solar PV Plant as well as the BESS incorporated with the PV Plant.

Contribution of Photovoltaic Power Systems to Frequency Control
Submission-ID 038
Julia Seidel, Bernd Engel
TU Braunschweig Institut für Hochspannungstechnik und Elektrische Energieanlagen - elenia, Germany
Today, the frequency of the grid voltage is predominantly controlled by conventional power plants with synchronous generators. However, the power system in Germany has been changing significantly over the past 20 years. More distributed energy resources (DER) have been integrated. The share of renewable energies in the gross electricity consumption increased to 36.2 % in 2017 and this trend is still persistent [1] . In a prospective 100 % renewable energies scenario the amount of synchronous generators will decline. Therefore it is essential that variable renewable energies take over tasks of frequency control to reduce the number of must-run units and to conserve the high security of supply. In this paper, one approach of involving PV systems in stabilizing grid frequency is presented.

The only requirement for PV systems concerning frequency control is the ‘frequency-depending active power reduction’. At 50.2 Hz, all DER are obliged to reduce their active power as a function of grid frequency. [2] In the normal operating state of the power system, where control reserve is applied for stabilizing grid frequency, PV systems are not involved. A participation in the control reserve market is hence not reasonable today.

In [3] , it is shown that PV systems are able to deliver control reserve. The main challenge, determining the maximum power point while operating below, can be achieved with high precision. The provision of Frequency Containment Reserve (FCR) with a PV inverter was successfully tested under laboratory conditions in [4] . Due to the characteristics of inverters these systems are able to react to a frequency deviation even faster than FCR is requesting. This advantage can be transformed into a new type of control reserve, called fast Frequency Containment Reserve (fFCR). This can help integrating inverter-coupled DER into frequency control mechanisms. At the same time, the decreasing inertia in the power system can be averted. In this paper, fFCR will be implemented as a new dimension of frequency control.

In order to develop a new measure of control reserve, it is first analyzed which parameters are most important to stabilize frequency. In a sensitivity analysis these parameters, e.g. dead band or gradients, are varied in order to define appropriate requirements for an fFCR delivered by inverter-coupled DER.

Afterwards, a system stability study is conducted in order to evaluate the impact of the fFCR on sudden frequency deviations. Laboratory measurements validate the results and provide a promising impact on the power system.


[1] Umweltbundesamt, Erneuerbare Energien in Deutschland, 2018 [2] VDE e.V., VDE-AR-N 4105 Erzeugungsanlagen am Niederspannungsnetz, Berlin, 2011 [3] B.Osterkamp et al., Photovoltaikanlagen in der Regelleistung, EW – Magazin für die Energiewirtschaft, 2017 [4] S. Laudahn et al. , Frequency Control in Distributed Inverter-Dominated Electrical Power Systems, Solar Integration Workshop, Vienna, 2016

Generator technical performance standards in the Australian National Electricity Market
Submission-ID 041
Julian Eggleston
AEMC, Australia
The electrical power system that makes up the National Electricity Market (NEM) in Australia is being transformed by the increased integration of asynchronous wind and solar generation, at the same time as synchronous generation is operating less or being retired.

The associated reduction in short circuit ratios at new generation connection points, and the different technical characteristics of the asynchronous generation, has resulted in the need to review the technical performance standards that connecting generators are required to meet. This review of the technical connection standards has also been informed by analysis of the black system event in South Australia in 2016.

The paper will include overviews of the NEM:

This presentation could focus on either the negotiation process, technical aspects of the access standards or a balance of the two.

The provision of primary frequency control in the Australian National Electricity Market
Submission-ID 042
Julian Eggleston, Ben Hiron
Australian Energy Market Commission (AEMC), Australia
The control of the frequencies of the Australian NEM mainland and Tasmania has been deteriorating in recent times. In particular, the distribution of the frequency within the normal operating frequency band (i.e. in the absence of a contingency such as the trip of generating unit) is significantly flatter than the normal distribution experienced when the NEM started in 1998. This issue was discussed at the 16thWind Integration Workshop in Berlin in 2017.[1]

Since then the system operator (AEMO) and the market rule maker (AEMC) have been reviewing the arrangements for managing the frequency within the normal operating frequency band. It has been confirmed that, as discussed at the 16thWind Integration Workshop, one of the drivers of the degradation of frequency during normal operation has been a reduction of primary frequency control provided by synchronous generation within the normal operating frequency band.

Under the current market rules generators are not required to vary their active power output to help correct frequency deviations unless they are enabled to provide a specific Frequency Control Ancillary Service (FCAS). The existing FCAS markets in the NEM do not include arrangements for the provision of primary frequency control within the normal operating frequency band. Therefore AEMO and the AEMC are considering new regulatory and market arrangements that will result in the sufficient provision of primary frequency control within the normal operating frequency band.

The paper will include:

[1] “Lessons learned from the Australian frequency control ancillary service market”, K. Summers, R. Jennings and J. Peters, 16thWind Integration Workshop, Berlin 2017.

How to Create Value through Aggregation: A Business Model Review for Multiple Regulatory Environments in Europe
Submission-ID 044
Simon De Clercq 1, Daniel Schwabeneder 2, Carlo Corinaldesi 2, Odilia Bertetti 1, Achim Woyte 1
1 3E sa, Kalkkaai 6 B-1000 Brussels, Belgium
2 Energy Economics Group, Vienna University of Technology, Gusshausstrasse 25-29/E370-3, 1040 Vienna, Austria
The rising share of distributed generation is having a profound impact on European electricity markets. Increased variability and price volatility require a more robust power system that allows alternative forms of production and consumption. Aggregation of renewable energy sources in a combined portfolio can significantly accelerate their market integration by diversifying revenue mechanisms, offering alternative ownership options and increasing asset valorisation. Today we see a variety of business models for aggregation, which are largely driven by the national regulatory framework and the distributed resources available for which aggregation can create value. This paper reviews three aggregator business models that were conceived by commercial aggregators in different regulatory environments in Europe. On the theoretical side, the business models are simulated using a business economic optimisation model. The practical implementation is monitored and the aggregators’ implementation experiences are discussed. The paper identifies and documents how the business models can create value for the aggregators, their customers and for society in the different regulatory environments.

Results from H2020 BestRES, EC GA 691689.

The Impact of PV on The Ability of Storage to Provide Peaking Capacity
Submission-ID 045
Paul Denholm, Robert Margolis
National Renewable Energy Laboratory, United States
As the cost of battery energy storage declines, there is increased opportunity for it to provide an alternative to conventional peaking resources such as gas turbines. This presents a very large technical potential, measuring in the hundreds of GW in the United States alone. For energy storage to achieve a significant market share, its costs must be reduced to the point where it has similar life-cycle costs to combustion turbines. This cost parity is much easier with short-duration (4 hours or less) batteries. However, batteries must demonstrate their ability to provide high capacity credit, to provide a true alternative to non-energy limited resources. The ability of energy storage to provide peak capacity is based largely on the shape of electricity demand during periods of highest demand, which correspond to periods of highest loss of load. In this work, we examine the ability of storage to reduce peak net demand in various regions of the United States. We demonstrate how short-duration storage (4 hours or less) may have a relatively small technical potential, as storage effectively clips and “widens” the peak net demand period beyond 4 hours. However, he addition of significant PV can change the shape of the net demand patterns, narrowing the peak and therefore increase the potential for shorter duration storage. We show how this potential is more significant in areas with strongly summer peaking demand and reduced in winter peaking systems with longer duration peaks.

Hosting capacity of the grid for photovoltaic installations – a stochastic approach applied to single-phase connections
Submission-ID 046
Math Bollen 1, Enock Mulenga 1, Sarah Rönnberg 1, Nicholas Etherden 2
1 Electric Power Engineering, Luleå University of Technology, Sweden
2 Vattenfall R&D Power Technology, Sweden
This paper discusses the voltage rise due to PV installations connected to a low-voltage network. The connection of individual installations is studies for both single-phase and three-phase connection. A stochastic method is presented to estimate the hosting capacity. This method is illustrated for random and coordinated connection of single-phase installations. It is shown, in a number of ways, that the installation of large (e.g. 6 kW) single-phase connected units can easily result in unacceptable overvoltages.

Investigation of Balancing Power Demand in the Roll-out Scenario of Solar Energy in Senegal
Submission-ID 053
Birgit Koeppen 1, Anna Kerstin Usbeck 1, Mamadou Lamine Ndiaye 2, Amadou Ndiaye 2
1 Hamburg University of Applied Sciences, Germany
2 Université Cheikh Anta Diop, Senegal
The total power demand in Senegal raised continuously over the last years. To face this increase, especially the use of solar power is assumed to be an efficient way as the solar irradiance near to the equator is high. But the connection of more and more photovoltaic parks to the Senegalese grid leads to an increasing demand of balancing power. The aim of this paper is to analyze the amount of this balancing power demand. For this aim the specific conditions of Senegal are analyzed based on measurement data of power demand and production as well as on measured climate data. Based on this the required balancing power for different scenarios is calculated. Considering the specific conditions in Senegal, a first analysis of possible concepts for providing this balancing power, especially by storages, is presented.

Interaction analysis of large-scale PV power plants considering the AC network
Submission-ID 055
Eduardo Prieto-Araujo, Ricard Ferrer-San-Jose, Oriol Gomis-Bellmunt, José Montero-Casinello
Large Photovoltaic (PV) power plants are systems dominated by power electronic converters, as each PV module is interconnected to the plant grid through an inverter. These systems are coordinated by a higher level Power Plant Controller (PPC). Based on the Transmission System Operator (TSO) set-points, the PPC sends active and reactive power references to each of the inverters to adequately operate the power plant. Such a system can present stability and interactions issues due to the large number of converters operating to the same network, posing a challenge on the power plant control design at all levels.

In order to design an integral plant control system that ensures an adequate system performance, the relevant system dynamics must be considered. First, the fast dynamics of the PV module inverters, which impose the system response at each of the production nodes. Also, the internal grid of the power plant, which would present a specific dynamic response based on the defined system layout. Additionally, other elements connected to the power plant grid such as passive reactive power compensation systems, energy storage systems or FACTS devices might also affect the plant dynamic response. Besides, the PPC imposes the higher level dynamics of the complete power plant from the grid perspective. Finally, the strength of the AC grid where the power plant is connected, can also impact the system dynamics. Therefore, considering that a PV power plant gathers the previously described dynamics, a thorough analysis must be performed to ensure an adequate system performance in order to avoid problems during the operation.

To do so, this article presents an integral power plant control design methodology able to mitigate eventual system interactions. First, a detailed dynamic model of the PV power plant is developed in detail, including the PV inverters, the internal network, the PPC, and the AC grid, among other elements. Then, a detailed dynamic analysis of the derived mathematical model is performed to identify the potential system interactions and the oscillatory modes of the system. Based on this analysis, a design of the different controllers involved in the power plant is performed in order to guarantee a proper system response. Finally, the theoretical results are validated through dynamic simulation results of the complete solar power plant.

Managing frequency in low inertia grids
Submission-ID 057
Scott Hagaman, Tim George, Jennifer Crisp
DIgSILENT Pacific Pty Ltd, Australia
Australia and other electricity grids are facing many issues with the increase of asynchronous generation commonly associated with renewable energy. One of the concerns is that renewables will displace synchronous plant and associated inertia.

Demand-supply imbalance is compensated by rotational energy stored using inertia. As inertia is reduced in the power system, the rate of change of frequency (RoCoF) following a system contingency or fault increases and at some point, the existing frequency standards cannot be maintained. Fast RoCoF has the potential to compromise the stability of the power system. Some reasons for this include:

  • Mechanical torque on rotors increases for rapid changes in frequency, which may cause damage or a turbine trip and this further exacerbates frequency control problems
  • Historical frequency control mechanisms such as load shedding and turbine governor control may be inadequate in low inertia systems due to their relatively slow speed of response in comparison to the rate of change of frequency

The transition from the current operating regime to low inertia power systems poses challenges and many utilities around the world including Australia are seeking to mandate minimum system inertia to ensure the grid continues to operate as it always has. Whilst this is one way to solve the problem, it incurs economic costs on all users of the power system and there are likely to be alternative solutions.

As inverter-based generation systems are able to rapidly increase or decrease active power, they could potentially be made to respond quickly to frequency variations. Faster frequency control plant may be able to provide sufficient ‘synthetic inertia’ or act as an alternative for inertia.

In terms of load response, fast communications systems could potentially be used to assist in load shedding when the detection of rapid rate of change of frequency is too slow. Inverter controlled load such as air conditioners, pumps and fans could also be programmed to respond to frequency variations and provide load relief that would otherwise not be available.

This paper investigates the operation of a low inertia grid to examine the impacts on the frequency standard, system operations and viable options to ensure continued operation of the interconnected power system as inertia is reduced.

The authors seek to investigate what would be required to operate a practical power system, consisting of a variable mix of new and old technology generation, with very low inertia compared with current levels. An interconnected 2 region grid will be configured to resemble a practical real-world example.

The objective of the paper is to facilitate economic generation deployment and grid operation by offering an alternative to mandatory rotational inertia levels provided by synchronous machine technology.

Japanese Power System Operation with Large Amounts of PV using Day-ahead and Intraday Unit Commitment
Submission-ID 059
Yusuke Udagawa 1, 2, Yuki Nishitsuji 1, 2, Kazuhiko Ogimoto 1, Joao Gari da Silva Fonseca Junior 1, Katsuyuki Ukegawa 2, Suguru Fukutome 3
1 The University of Tokyo, Japan
3 JP Business Service Corporation, Japan
The amount of installed and approved photovoltaic (PV)s reached 28.5 GW and 84.5 GW, respectively, in march 2017 in Japan under Feed in Tariff Program, which started in 2012. In the near future, about 16 GW of PV are expected to be installed only in the Kyushu island area, which has a peak demand of about 15 GW in summer. Under these conditions, the power system operation in Kyushu is expected to face the need for curtailment of PV. To carry out appropriate PV curtailment, it is required the use of weather forecasts with unit commitment planning (UC) and also recommitment with high frequency of updated forecasts. With this background as motivation, we analyzed the operational situation and the operational cost (fuel cost and start-up cost) of the power system of Kyushu area in Japan in 2030, considering high penetration of PV and need for curtailment of PV. In this study, we calculate the effect of multiple weather forecasts (long term and short term) and its updating against the power system operation and curtailment of PV using unit commitment model, with re-commitment process. The re-commitment process reflects improvement of the accuracy of PV yield forecasts according to shorter forecast horizons. In this way, our day-ahead and intraday UC model can take into account PV yield forecasts and errors. The PV yield forecast and its error were updated in every 30 minutes in the intraday UC to carry out the recommitment. The results of intraday UC allowed us to evaluate the value of forecasts according to different indicators such as accommodation of variable renewable energy (the amount of required PV curtailment), economy (operational costs of day-ahead and intraday forecasts), and security of supply (Expected Unserved Energy or the generation of the last resort generator).

PV integration with flexible generation and consumption units – Evaluation of a quota-based grid traffic light approach in a field test
Submission-ID 069
Kilian Geschermann, Katharina Volk, Christian Lakenbrink, Martin Konermann
Netze BW GmbH, Germany
The increase in renewable power generation in Germany leads to the need for new concepts for the coordination and utilization of flexibility provided by decentral units for congestion management in distribution grids. One approach is the grid traffic light concept. This paper introduces the quota model as an implementation of the yellow traffic light phase. Within this model, quotas represent the share of flexible units in a grid cluster that can be activated at the same time without causing congestion. Based on results of a field test in a rural low voltage grid, it is shown that grid congestion caused by PV feed-in can be successfully avoided by the applied quota based approach. For this, the flexibility of battery storage units is used. By storing feed-in peaks, the curtailment of PV generation can be avoided and a higher share of renewable power generation can be integrated into the grid. In addition, the application of the quota model puts the distribution system operator in a position to bridge the time gap until grid expansion is carried out or to possibly postpone or avoid grid expansion if congestion occurs only in a few hours per year.

Integration of PV+storage - technical and economic evaluation at distribution grids
Submission-ID 072
Guangya Yang 1, Peng Hou 1, Dezso Sera 2, João Pedro Rodrigues Martins 2, Philip Douglass 3, Sebastian Martens 3, Poul Mose Johansen 4, Daniel Karnøe Svendsen 4, Klaus Moth 5
1 Technical University of Denmark, Denmark
2 Aalborg University, Denmark
3 Dansk Energi, Denmark
4 Integrate, Denmark
5 LivingPower, Denmark
Incorporation of energy storage units to solar PV plants in small-scale applications has been attracting major attention from the manufacturers worldwide in the past few years. This consolidated into both technology development and commercialization. Most of the solar PV inverter manufacturers nowadays have adopted battery storage as part of their product portfolio offered together with their solar inverters. Some also provide standalone battery storage units to the consumers that can directly cooperate with the existing solar PV plants and/or loads. Despite short-term regulatory risk, the application seems still viable in the long term, where the key business case lies in the reduced cost of electricity bills and the ever-decreasing capital costs of solar PV and storage. In due course, distribution grid operators have become concerned about the possible reduced revenue from the existing tariff model due to the increasing level of self-sufficiency of the consumers. The paper summarizes the background, ideas and the initial results from the work in a Danish project whose activities cover both the technology development and operational analysis of PV and storage systems.

Towards improved and more scalable PV power forecasting: Insights from Solcast’s application of third-generation geostationary satellites, PV power measurements and low-cost sky imagers
Submission-ID 073
James Luffman, Harry Jack, Darren Reid
Solcast, Australia
Solar forecasting technology has been challenged in recent years by the rapid rise in solar PV penetration globally. Initial bespoke solar forecasting solutions for utilities with high solar penetration have often not delivered on user expectations of accuracy, and have often been quite 'heavy' in terms of cost, timeline and the human effort to configure and use. The number of user applications for solar forecasting is increasing exponentially due to increasing solar, and also due to the rise of complementary and enabling technologies such as paired storage, virtual power plants and hybrid or off-grid power systems. Even within a single utility, multiple automated systems and databases need to consume solar forecasting information. As the number of users and their applications increases, bespoke solutions can no longer deliver on the full set of applications, and the onus for solar forecasting solutions moves from the specific to the general. Secondly, the demand for improved forecast skill spreads along the lead-time horizon, rather than there being a lead-time of primary importance (e.g. day ahead or hour ahead) that can be tuned in a given application.

In this paper, we present Solcast's unique approach the dual challenge of accuracy and scalability. Firstly, data from the new third generation geostationary satellites is fused with dynamical model information to detect, track and forecast cloud cover. This is crucial for achieving useful skill in tropical and sub-tropical applications, for dealing with cloud growth and decay, and for projecting satellite information down to minute timescales in forecasting applications. Secondly, we present the Solcast API - a massive data processing, storage and distribution system that enables the full range of users to instantly test and subscribe to high quality forecast and irradiance data, compare to local measurements, post their data for improved accuracy, and visualise forecasts and recent accuracy. A set of real applications of the Solcast API are presented as examples, across multiple continents, and across the forecast lead time range.

Modelling City Scale Spatio-temporal Solar Energy Generation and Electric Vehicle Charging Load
Submission-ID 076
David Lingfors 1, Mahmoud Shepero 1, Clara Good 2, Jamie Bright 3, Joakim Widén 1, Tobias Boström 2, Joakim Munkhammar 1
1 Uppsala University, Sweden
2 UiT The Arctic University of Norway, Norway
3 ANU College of Science, Australia
This study presents a model for estimating building-applied photovoltaic (PV) energy yield and electric vehicle (EV) charging temporally over time and spatially on a city scale. The model enables transient assessment of the synergy between EV and PV, thus is called the EV-PV Synergy Model. Spatio-temporal data on solar irradiance is used in combination with Light Detection and Ranging (LiDAR) data to generate realistic spatio-temporal solar power generation profiles. The spatio-temporal EV charging profiles are produced with a stochastic Markov chain model trained on a large Swedish data set of travel patterns combined with OpenStreetMap (OSM) for deterministically identifying parking spaces in cities.

The modelled estimates of solar power generation and EV charging are combined to determine the magnitude and correlation between PV power generation and EV charging over time on city scale for Uppsala, Sweden. Two months (January and July) were simulated to represent Sweden’s climate extremes. The EV penetration level was assumed to be 100% and all the roofs with yearly irradiation higher than 1000 kWh/m2 were assumed to have PV panels.

The results showed that, even in January with the lowest solar power generation and maximum EV load, there can be a positive net-generation (defined as the integration of PV generation minus EV charging load over time) in some locations within the city. Central locations exhibited a positive temporal correlation between EV charging load and PV generation.

Negative temporal correlations were observed in the outskirts of the city, where typically night time home-charging was prevalent. In the highest PV power generation month (July) the solar generation was 16 times higher than the EV charging load. Spatially, the net-generation was positive in almost the entire city. However, the time-series correlation between the EV charging load and the PV generation reached more extreme positive and negative values in comparison with January. This was a result of the higher variability in irradiance during July in comparison with January. In summary, we find that there is a favorable synergy of EV-PV technology within the city center with assumptions of workplace charging behaviors for both winter and summer months. An unfavorable synergy with suburban areas where typically nighttime charging behaviors negatively correlate to PV generation. This suggests that distributed PV should be targeted around city center/workplace EV charging stations.

Climate change mitigation potentials of vertical building integrated photovoltaic
Submission-ID 077
Tobias Blanke 1, Bernd Prof. Dr. Döring 2, Marius Vontein 3, Markus Prof. Dr. Kunhenne 3
1 RWTH Aachen University, Germany
2 FH Aachen Institute for Building technology, Germany
3 RWTH Aachen University Sustainable Metal Building Envelopes, Germany
Buildings caused approximately 19% of the energy related greenhouse gas (GHG) emissions in 2013 [1]. One option to reduce these emissions are Building Integrated PhotoVoltaics (BIPV) [1]. Therefore, this paper investigates the potential exceptionally of vertical BIPV within the different Shared Socioeconomic Pathways (SSP) [2]. Therefore, a carbon budget for the electrical energy consumption of the EU building stock is assumed, to compare these with the cumulated emissions of the electrical energy consumption of the EU building stock within the different SSPs. This is done with and without the massive use of BIPV.

The cumulated emissions are calculated for all EU countries and different BIPV cell types. These types are multi- (mult-Si) and mono-crystalline Silicon solar cells (mono-Si) as well as Thin-Film Photovoltaic (TFPV) like Copper Indium Gallium Selenide solar cells (CIGS), amorphous Silicon solar cells (a-Si) and Cadmium Telluride solar cells (CdTe). The solar irradiation is determined and life-cycle emissions of the different BIPV are taken from papers [3], [4], [5]. Besides, the data of the current EU building stock are taken from different sources [6], [7], [8] and are forecasted for the electrical energy system with the SSP scenarios [2], [9] and for the building stock with the scenario developed by ENTRANZE [10].

Implementation of the European Network Code on Requirements for Generators on the European national level - Trends and challenges
Submission-ID 082
Roland Bründlinger 1, Gunter Arnold 2, Giorgio Graditi 3, Nils Schäfer 2, Thomas Schaupp 4, Giovanna Adinolfi 3
1 AIT Austrian Institute of Technology, Vienna, Austria, Austria
2 Fraunhofer IEE, Kassel, Germany, Germany
3 ENEA - Italian National Agency for New Technologies, Energy and Sustainable Economic Development, Naples, Italy
4 KACO New Energy GmbH, Neckarsulm, Germany
The publication of the Commission Regulation (EU) 2016/631 establishing a network code on requirements for grid connection of generators ("NC RfG") will fundamentally impact the landscape of European national grid codes and standards governing the connection of generation from RES as well as non-RES. The NC RfG establishes a EU wide harmonization of grid interconnection requirements aiming at ensuring and increasing the system security with a growing share of variable RES as well as boosting the market of generation technology and increase competitiveness across Europe.

The individual requirements in the NC RfG can be categorized as:

  • Exhaustive requirements define capabilities of generators by function or principle and by specified parameters.
  • Non-exhaustive requirements define basic capabilities, without further specification of parameters and/or settings. The details have to be defined on the national level to take into account national or regional system characteristics
  • Mandatory and Non-mandatory requirements, which must be met by the generator resp where the relevant system operator can decide upon their application

According to its nature as European Regulation, the NC RfG will be "binding in its entirety and directly applicable", with the invidual requrirements being applied from April 27, 2019, 3 years after publication. Until this date, existing national specifications, codes, standards and other documents have to be in-line with the NC RfG. During the first two years of this period (until May 2018), national specifications have to be reviewed, amended or adapted to meet the provisions, including detailling non-exhaustive requirements. The remaining year will allow for notification of the documents and ensure product compliance. As part of the national implementation, also the maximum capacity thresholds for the generator types specified have to be set by the relevant authorities.

The paper will provide an overview on the status and details of the NC RfG implementation in the main national markets for PV in Europe. For specific, non-exhaustive requirements of the RfG which are relevant for grid connected PV, e.g. related to FRT, reactive power, control and monitoring functions, certification etc. a detailed comparison of the national specifications will be given. The overview on the national implementation of the NC RfG requirements will be complemented by a summary of the EN 50549-X series of standards setting the requirements for generating plants to be connected in parallel with distribution networks.

Evaluation of power flow prognosis methods for congestion management in low voltage grids
Submission-ID 085
Katharina Volk 1, Christian Lakenbrink 1, Nina Hatje 1, Peter Stolle 2, Filippos Sivorotka 3, Kevin Förderer 4
1 Netze BW GmbH, Germany
2 Fichtner IT Consulting AG, Germany
3 Seven2one Informationssysteme GmbH, Germany
4 FZI Forschungszentrum Informatik, Germany
For the application of proactive congestion management concepts the enhancement and implementation of power flow forecast methods is essential. The increase of directly marketed renewable generation and the use of flexibility options by market participants may lead to difficulties in predicting the power flow without any knowledge of market actions. Therefore, power flow forecast methods that take data and information from market parties and local building energy management systems into account have to be developed and evaluated. This Paper presents an integrated process for creating power flow forecasts in low voltage grids based on the forecasts of different role-specific system solutions such as building energy management systems and flexibility management systems. Moreover, the use of standard load profiles and Smart Meter measurements for the creation of load profiles is presented. The results of the evaluation of the process and the methods for one low voltage grid in the gridlab Freiamt of Netze BW show the potential of the proposed process and the developed methods for a further congestion management in low voltage grids. Furthermore, possible measures to improve the methods are mentioned.

Identification of a Dynamic Equivalent of an Active Distribution Network from Monte-Carlo Simulations
Submission-ID 086
Gilles Chaspierre 1, Guillaume Denis 2, Patrick Panciatici 2, Thierry Van Cutsem 3
1 Dept. of Elec. Eng. and Comp. Science, University of Liège, Belgium
2 Research & Development Dept. RTE, France
3 Fund for Scientific Research (FNRS) University of Liège, Belgium
Context and motivation

Distribution grids are expected to host more and more distributed generators connected to the grid through an electronic interface. This will completely change the dynamics of distribution grids in a near future. It becomes urgent to have proper models of such Active Distribution Networks (ADNs) in order to account for their contributions in power system dynamic studies.

For transmission system operators, a unique model of both transmission and distribution systems is not only impractical, but would also entail data confidentiality issues.

It makes sense for the Distribution System Operators (DSOs) to process the data of their own systems and transmit to the TSO reduced, ``anonymized'' models of significantly lower complexity than the original, unreduced model they have assembled.

For the identification of the ADN equivalent, the “simulation-based” approach has been preferred, where the ADN responses are simulated, and model parameters are adjusted to fit those responses. Among this family of method, the “grey-box” approaches have been retained in this work. They rely on a reduced model that mimics the behaviour of the components it replaces, leading to a model with known structure but unknown parameters.

For DSOs, one major issue when setting up a detailed ADN model lies in the uncertainty affecting the behaviour of its components. In this work, it is assumed that the individual dynamic behaviour of Inverter-Based Generators (IBGs), loads, etc. can be reasonably well captured by a parameterized model, but the values of its parameters are uncertain.

A well-known approach to deal with such uncertainty consists of performing Monte-Carlo (MC) simulations, involving in this case random variations of the parameters. Thus, for given disturbances and operating points, a set of randomized time responses is generated.

The next step is to extract from this large set some confidence intervals, i.e. time varying ranges of values in which the dynamic responses of the detailed ADN model can lie.

Extracting the confidence intervals

MC simulations generate instances of the ADN model. Typical disturbances are applied to each of them and the corresponding active and reactive power responses are collected.

The next step consists in computing the percentiles 5 % and 95 % of the dynamic power responses. The latter make the confidence intervals in which 90 % of the dynamic responses should lie.

Identification of the ADN equivalent

The parameter of the ADN equivalent are adjusted until the responses lie in the confidence intervals extracted from the MC simulations of the original, unreduced system.

At this stage of our research, a meta-heuristic optimization method has been used, more precisely a variant of the Differential Evolution algorithm.

Design of a Hybrid-Power-System dominated by solar power and battery storage including retro-fitting options driven by future load increase
Submission-ID 089
Meike Kühnel 1, Benedikt Hanke 1, Olga Weigel 2, Ingo Stuermer 3, Alister McMaster 4, Sander Maebe 2, Karsten von Maydell 1
1 DLR Institute of Networked Energy Systems, Germany
2 Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH, Germany
3 Lower Saxony Ministry for Environmental Affairs, Germany
4 Department of Economic Development, Environmental Affairs and Tourism of Eastern Cape, South Africa
Energy use and the access to reliable electric power is an essential cause and facilitator of economic grows and a binding constrain for all sizes of business . With 15.8% of South Africa’s population without access to electricity and poverty head count ratio of more than 20% reducing the backlog in electrification is one aspect in the anti-poverty campaign of the country. In far off rural areas islanded mini-grids based on renewable energy generation offer both, access to electricity on a grid-like level and an economic solution for the responsible authorities. The main challenges for planner, developer and engineers of mini-grids is the appraisal of a residential load profile of former un-electrified areas and the matching with highly fluctuating renewable energy resources. The adjacent increase of the load caused by economic grows, new demands for electric appliances and a growing village size triggered by the availability of electricity adds further challenges and uncertainties. We developed a hybrid pv-diesel mini-grid with the ability to supply the first year electric power de-mand with a power mix including 86% to 100 % renewable energy by a pv and battery storage. The mini-grid is designed to connect 55 households with a mean energy consumption of 50kWh to 150kWh per month and household depending on the predicted load scenario in the first year of operation. The annual power consumption of the village is expected to increase by a factor of 1.5 to 2.4 during the first 5 years. Within 15 years the energy demand could be over 4 times higher than the first year demand the system is designed for. With growing demand the diesel consumption increases, changing the energy mixed from mainly renewable to a fossil fuel based grid. The cost of fuel becomes the main driving factor for the levelized cost of electricity (LCOE). For a five year project lifetime the cost of fuel is responsible for 29% to 71% of the LCOE depending on the load scenario. The economic decision of energy provision by hybrid mini grid can be a financial risk if future load developments are included. Including retro-fitting structures and a growing renewable power generation plant into upfront consid-erations are essential for an economical sustainable and persistent mini-grid. Eduction and higher awareness of energy consumption and efficiency can only insufficiently reduce this aspects. An adaptive system will learn from the development of the villages’ electric power consumption during the first years of electrification. A recommendation for further installation of renewable energy capacity such as micro wind turbines and additional solar power will be triggered by the recorded demand development and additional onside data using the infrastructure of the generation side. This opens up a new look on business plans and the calculation of LCOE with a more fluently and dynamic behavior reflecting real life challenges.

Agent-based optimization of retail electricity rates for PV integration
Submission-ID 091
Michael Hinterstocker, Serafin von Roon
FfE GmbH, Germany
Rising shares of PV energy in the energy system causes increasing necessity for congestion management in order to maintain the stability of the electricity grid. The obvious way of dealing with problematically high PV generation is curtailment of PV plants in the affected area, but there are other flexibility options that can be utilized for balancing generation and consumption. Residential demand-side management is one of these that has been widely discussed for several applications and potentially makes a substantial contribution to the avoidance of curtailment by load-shifting to periods with high PV generation. Participation of residential customers has to be properly incentivized in order to tap this potential. This can be achieved by the implementation of variable electricity rates that encourage increased consumption in these periods of time. Therefore, these rates are to be designed in a way that is both appealing to the potential customer and at the same time representative for the grid state.

There is a vast variety of rate structures which can be characterized by a number of parameters, which for example define the energy price at different times of the day. More sophisticated structures might include demand charges, dependence on e. g. season, grid state, or renewable generation, as well as features like interruptibility. In order to determine these parameters, an agent-based optimization approach is proposed.

Based on measured consumption of residential customers in high temporal resolution, their behavior and consequentially their potential for load-shifting is analyzed via pattern recognition algorithms. In combination with parameters that define individual preferences regarding acceptance, expected savings, suitable appliances etc., this allows simulating the reaction of individual households to given rates.

The reaction to given rates not only includes the expected load-shifting measures, but also the switching decision to one of the other offered rates. Therefore, after several time steps, the agent-based simulation yields the market penetration of each considered rate, and the resulting effects on the individual customer as well as the overall grid situation. By defining the reduction of PV curtailment as optimization objective, this result can be used as an input for an optimization algorithm, which determines the parameters in the optimal way.

Thus, the described simulation and optimization model yields the optimal parameter set for variable retail electricity rates that incentivize the avoidance of curtailment. Plausibility checks show that the model works as expected. Two main conclusions can be drawn from the results: On the one hand, complex rate structures achieve only very low market penetration and have therefore low impact on the system, and on the other hand, simple time-of-use structures are not capable of representing the grid requirements. Therefore, the optimization helps to find the right balance.

Submission-ID 094
Andrew Kitimbo, Ulrika Morild
Vattenfall AB, Sweden
In the past couple of years, the world has experienced an increase in the amount of electricity generated from solar PV and wind. Furthermore, developments in solar PV generation technologies have allowed solar PV to be modular and affordable, thereby enabling its installation on residential and commercial buildings spread out all over the low voltage distribution grid.

However, the traditional electricity distribution grid is designed to transfer power from centralised generation sources to loads downstream in the grid, implying that the various grid components and grid services such as reactive power compensation, are dimensioned to support a grid where power flow is from the traditional central source to a known or expected amount of load.

With increasing penetration of solar PV in the low voltage distribution grid, there is a possibility that the PV production outweighs the local power demand in many locations. This would result in the power generated being transferred upstream through the distribution grid, giving rise to reversed power flow in the grid.

The flow of power in the distribution grid in a direction contrary to that for which the grid was designed presents the risk of possible over voltages which could occur due to the shifts in the flow of reactive power in the grid (contrary to the designed flow) and possible overloading of grid components such as distribution transformers and lines.

Vattenfall is conducting an investigation into the use of battery energy storage in the distribution grid to mitigate the risks presented by increased penetration of solar PV. The investigation is being done with a physical battery energy storage system installed in a real distribution grid with several households and a significantly large solar PV installed capacity. In this study, the impact of the battery storage on the voltage profiles at various nodes in the grid and the loading of selected grid components shall be investigated and presented in this paper.

Towards dual photovoltaic and thermal self consumption
Submission-ID 098
Aziz Naamane
Aix-Marseille Université, Université de Toulon, CNRS, LIS, SASV, Marseille, France, France
The speech on the self consumption generally concerns photovoltaic, but we forget too often the thermal consumption Or, with the modules PV/T The self consumption is double because these modules produce both photovoltaic electricity and hot thermal water. Hot water produced by PV/T solar panels is inherently self consumed. The domestic hot water balloon is an natural way of storage of the system, and the hot water produced is thus 100% self consumed. The objective of this paper is to study in the first time new hybrid photovoltaic and thermal panels of inexpensive with new structures and new material for the absorber and proceed thereafter to the correct Sizing to perfectly meet the needs of the house and maximizing the energy produced in sync with the energy consumed. These solar systems (thermal and electrical) in urban and suburban areas. The main idea is to prove that PV/T systems can certainly satisfy the thermal levels required for low operating temperatures (e.g. indoor hot water (ECS), heating, pool water heating etc...)

The following parts will be presented in this paper:

• State of the art on PV/T sensors and methods for evaluating their thermal and electrical performance

• Hybrid sensor modeling and Matlab simulation

• Choice of an appropriate and applicable method for evaluating collector performance

• Presentation of the test facility to test the thermal and electrical performance of the PV/T sensor

• Evaluation of test results

• Simulation under TRNSYS of the overall system of hot water production

Should waveform distortion (harmonics) be considered when connecting large amounts of solar PV?
Submission-ID 099
Sarah Rönnberg, Tatiano Busatto, Math Bollen
Electric Power Engineering, Luleå University of Technology, Sweden
This paper aims at putting harmonic studies in low-voltage network on the agenda, especially with the increasing amounts of PV connected to such networks. A general overview is given of emission from PV inverters, where emission at “classical harmonics” is lower that other low-voltage installations but at “non-classical harmonics” it is higher. Preliminary results are shown for a measurement campaign mapping the margin available before harmonic limits are reached. It is shown that more than a certain fraction of customers with PV will reduce the resonant frequency of the low-voltage grid. The paper concludes with a discussion on how to approach the harmonic issue at low voltage with PV.

Optimal design of charging stations for all battery electric city buses
Submission-ID 104
Hany Farag 1, Nader El-Taweel 1, Moataz Mahmoud 2
1 York University, Canada
2 McMaster University, Canada
Both transportation and utility sectors retain various challenges to facilitate seamless integration for Battery Electric Bus (BEB) fleet systems; these are mainly on the spectrum of achieving convenient transportation systems while preventing their negative impacts on the utility grid. Among those challenges is the lack of appropriate simulation tools to model, design, and optimize BEB fleet systems. Accordingly, this paper proposes a new mathematical formulation to model BEB fleet systems. The model considers the operation philosophy of public bus transit systems and the energy consumption characteristics of BEBs. The proposed transit model is then integrated with the power distribution networks model to develop a problem formulation for the optimal design of BEB systems. The formulated optimization problem aims at determining the configuration of BEB fleet systems including: the bus battery capacity, chargers rated power, and the total number of installed chargers in the bus charging station. The integrated utility-transit model is tested using real-world transit system data to validate its efficacy.

The potential of using residential PV-battery systems to provide primary frequency control on a national level
Submission-ID 110
Rasmus Luthander, Samuel Forsberg
Uppsala University, Sweden
To keep the frequency stable in a synchronous electric grid, power sources providing primary frequency control (PFC) are needed. Today, hydro power dominates the PFC market in the Nordic countries. However, if PFC can be offered as an ancillary service from PV-battery systems, the hydro power could be used more efficiently for low-cost electricity production. This can also improve the profitability of PV-battery systems. In this study of 2231 detached houses in Sweden, the potential to use residential PV-battery systems for PFC is examined. The results show that prices of up to 500 EUR/kWh excluding VAT for a 2.5 kW/5 kWh battery storage system can make enough PV-battery systems profitable to replace existing PFC sources in Sweden. The results are based on hourly electricity and PFC prices from 2015–2017 and a discount rate of 0%. If the prices for PFC are reduced by 50% and the discount rate is 5%, a system price of roughly 100 EUR/kWh excluding VAT would be enough to reach the same goal. The battery storage needs to be replaced during the lifespan of the PV system, meaning that the battery system prices are the average of the expected prices during the coming 25 years.

Balancing Power Capacity Analysis for Primary/Secondary/Tertiary Controls Based on AGC30 Model with PV Penetration
Submission-ID 111
Bo Jie 1, Takao Tsuji 1, Kenko Uchida 2
1 Yokohama National University, Japan
2 Waseda University, Japan
Recently, because of the environmental burden, natural hazard, and energy dilemma in Japan, renewable energy, especially PV and wind power, become a key technology in the development of power systems. Here, because of the uncertainty of PV output depending on the weather conditions, it became an important issue to keep supply and demand balance with proper frequency due to the uncertain output fluctuation. In order to overcome the problem, various countermeasures have been widely studied and developed so far.

In 2016, IEEJ developed a new simulation model for frequency regulation, AGC30, which can simulate the supply/demand balance control considering both economic dispatching control (EDC) and load frequency control (LFC). On the other hand, in Japan, the balancing power market will be established in 2020, for ensuring the supply/demand balance in Japanese power system. It is an important issue to apply the appropriate auction mechanism in the balancing power market. To this end, authors have focused on the clearing method of the market and analyzed the differences between uniform single price and pay as bid auctions by using multi-agent simulation in which the bidding strategy of each supplier was determined by Q-Learning.

Solar and Wind Energy Combination
Submission-ID 114
Uppsala University, Sweden

The deployment of Renewable Energy Sources is driving power systems towards an important green transition. Many people think wind power and solar power are competitors where in fact they can complement each other. In this regard, there is a need to develop models to accurately capture the variability of more than one source of energy at a time. The idea in this case is to try to combine Wind and Solar to see if there are any major benefits in sharing the electrical grid connection already paid for, for the existing wind turbines. This analysis is based on a case study of ten locations in eight cities in Sweden with different geographical and temporal scales: Gotland, Övertorneå, Vadstena, Simrishamn, Aneby, Kommerberget Berg’s kommun, Sölvesborg, Kungsbacka. The methodology consists of studying solar energy production based on solar panels characteristics and time series solar irradiation derived from the Swedish Meteorological and Hydrological institute (SMHI) compared to wind turbines energy production for hourly data from 2015 to 2017. This combination is coupled to an energy storage system that takes the excess in times where production is higher than consumption and delivers it back to the grid in times where the production is lower. Finally, the analysis is followed by a financial study of the system costs and profits based on renewable current market prices. The results show positive profit indexes of the plan including the storage system. This mix is not the ultimate solution for covering the energy needs but it is a step forward into increasing renewables portion to satisfy both the energy business and the environment.

A Hybrid Solar Radiation Forecasting Based on Data Mining techniques and SVR
Submission-ID 115
Rakesh Kumar, Vivek vijay
Indian Institute of technology, India

The integration of renewable energy into an electrical network cause power quality and stability issues. Solar forecasts on multiple time horizons become increasingly important for the effective operation of the power grid, load following production, power scheduling and unit commitment. This paper proposes a novel hybrid solar radiation forecasting framework using seasonal and trend decomposition using loess (STL), discrete wavelet transform (DWT) and Support Vector Regression (SVR). The forecast error of the proposed method is calculated using metrics, RMSE and MAE. The performance of the proposed model is compared with other competent models present in the literature. Comparative analysis with the benchmark model shows better forecasting results for the proposed model.


In this research, we propose a novel hybrid framework based on data mining technique which uses both STL and wavelet transform, for hourly solar irradiance forecasting. Support Vector Regression (SVR) is used as a predictor. Data mining through decompositions helps in better characterization of global horizontal irradiance behavior and provides more appropriate learning of neural network to enhance the accuracy of forecast results. We use STL to decompose solar irradiance data into seasonal, trend and remainder components. Residual component of the data is then obtained by subtracting seasonal component from the data, which is also the sum of trend and remainder components. Residual component significantly contributes to changing dynamics of the solar time series data, whereas the seasonal component shows the day to day repetitions of the data series and so it does not contribute to intermittency. The residual series is further decomposed by using wavelet transform. The wavelet transforms act as a pre-processor and decomposed the time series into several sub-series with more detailed periodic information, which is easier to predict. Each decomposed sub-series of the data is then used for forecasting by SVR. The final forecast is obtained by adding SVR results and the previously obtained seasonal data series.

A Comparative Assessment of the Small Scale Distributed Generation Policies in the EU and Latvia
Submission-ID 117
Laila Zemite, Lubov Petrichenko, Antans Sauhats, Olegs Linkevics, Galina Bockarjova
Riga Technical University, Latvia
The focus of the paper lies on the feasibility evaluation of solar power generation technologies for the power system to achieve the balance of energy generation and support costs in a more efficient way. The case study is based on a time series of examples with an hourly resolution for several years of electricity prices from the Nord Pool power market, power demand and solar generation of Latvian prosumers. Electric energy generation and consumption data from more than 200 clients were collected and used. The article analyses the influence of net metering supporting scheme variants on the profitability of a projected investment of PV batteries. A recommendation to change the current net metering system to a more perfect one, taking into account the market cost of energy, concludes the paper.

Demand Side Management and Energy storage options for building’s PV integration.
Submission-ID 120
Zuleika Hoyos Cruz, Pedro Soares Moura
Institute of Systems and Robotics. Department of Electrical and Computer Engineering, Faculty of Sciences and Technology. University of Coimbra. Coimbra Portugal, Portugal
Nowadays, due to the climate and energy targets and the decreasing cost of the technology, the immersion of renewable energy sources, such as photovoltaic (PV) solar generation systems, is increasing around the world, bringing benefits for the environment, society and economy. Nonetheless, in residential buildings, PV generation and electricity consumption do not have the same variation profile and such mismatch brings the need to export to the grid a substantial quantity of the locally generated energy, even though an identical quantity of energy is later imported back for local consumption, leading to stress on the electricity distribution grid. Demand-Side Management (DSM) in residential buildings with PV generation along with battery storage can increase the potential of self-consumption and therefore the ability to increase the self-sufficiency of the building. The aim of this paper is to analyze DSM and energy storage options in order to evaluate the generated impact at the PV integration. The results show that by applying a rescheduling technique it is possible to decrease between 19.7% and 85.9% the energy that it is sent into the grid and, at the same time, it generates a reduction from 6.1% to 12.7% of the energy that it is consumed from the grid in order to ensure a more efficient matching between the generation power and demand power. By coupling a PV system to an energy storage system, the energy that it is injected into the grid, decreases between 45.1% and 100% and also does the percentage of consumed energy from the grid with a reduction between 7.1% and 50.4%.

Improving Demand Response techniques for migration to a smart micro-grid
Submission-ID 130
Nithin Isaac
Nithin Isaac, South Africa
Demand Response (DR) techniques and initiatives are employed to manage electricity consumption and cost of buildings, as well as to improve the reliability of the electric grid. With the transition to smart micro-grids, opportunities for DR are improving. The Council for Scientific and Industrial Research (CSIR) has introduced an Energy Autonomous program intended to make the campus energy autonomous by the year 2023. DR initiatives and techniques are seen to be one of the key influencers through which the CSIR campus can achieve up to 20% reductions in energy costs. There is requirement for good coverage of control strategies for different technologies. Currently, there seems to be limited strategies in place for DR initiatives such as lighting , air-conditioning and ventilation, heating loads, and other miscellaneous end-use systems. It is envisioned that improving current DR control strategies will allow the campus to better match end-user demand to supply, especially since benefits of DR include improved operational efficiency, particularly in systems with large renewables integrated. Additionally, it will reduce overall capital and operational costs allowing more funds to be available to fast-track the migration of the current grid into a smart micro-grid. This study focuses on areas for improvement within exisitng DR techniques and control strategies, in order to maximise energy reductions within the CSIR campus, and thereby, fast-track the migration of the current grid into a smart micro-grid.

Using solar and load predictions in battery scheduling at the residential level
Submission-ID 137
Richard Bean 1, Hina Khan 1, 2
1 Redback Technologies, 80-120 Meiers Rd, Indooroopilly 4068, Queensland, Australia
2 School of Information Technology and Electrical Engineering, The University of Queensland, 4072, Queensland, Australia
Smart solar inverters can be used to store, monitor and manage a home’s solar energy. We describe a smart solar inverter system with battery which can either operate in an automatic mode or receive commands over a network to charge and discharge at a given rate. In order to make battery storage financially viable and advantageous to the consumers, effective battery scheduling algorithms can be employed. Particularly, when time-of-use tariffs are in effect in the region of the inverter, it is possible in some cases to schedule the battery to save money for the individual customer, compared to the “automatic” mode.

Hence, this paper presents and evaluates the performance of a novel battery scheduling algorithm for residential consumers of solar energy. The proposed battery scheduling algorithm optimizes the cost of electricity over next 24 hours for residential consumers. The cost minimization is realized by controlling the charging/discharging of battery storage system based on the predictions for load and solar power generation values. The scheduling problem is formulated as a linear programming problem. We performed computer simulations over 83 inverters using several months of hourly load and PV data. The simulation results indicate that key factors affecting the viability of optimization are the tariffs and the PV to Load ratio at each inverter.

Depending on the tariff, savings of between 1% and 10% can be expected over the automatic approach. The prediction approach used in this paper is also shown to out-perform basic persistence forecasting approaches. We have also examined the approaches for improving the prediction accuracy and optimization effectiveness.

Verification of Difference in Voltage Ride Through Assist Performance of Voltage Management Equipment between Serial and Parallel Type by Experiment
Submission-ID 139
Akira Moriwaki, Satoshi Uemura
Central Research Institute of Electric Power Industry, Japan
In recent years, the penetration of renewable energy such as photovoltaics (PV) has been progressed in Japanese distribution system. The government promoted assistance to install PV by means of the purchasing surplus electricity system and feed-in tariff (FIT) system since 2009. Under the feed-in tariff 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 is guaranteed for 10 years from the start of the program (less than 10 kW). The consumer expired this program will appear since 2019. Therefore, the government promotes Microgrids such as self-consumption of PV in local area.

Case of PV centered Microgrids, fault ride through (FRT) is important because it makes a strong contribution to a stable supply of power until a network switches, a grid fault and so on. FRT is the function of power conditioning sub system (PCS) that continue to operation under power outage and instantaneous voltage drop. When a PCS doesn’t mount FRT, the relay in PCS is operated by rapidly changing of interconnected voltage and frequency and the PCS shutdowns. Other PCS are caused shutdowns by another one and this grids proceeds blackout. After bring the voltage and frequency back to normal, it takes PCS a while to restart. On the other hands, the PCS continues to operation and prevents blackout when a PCS mounts FRT. In more details, PCS interconnected high voltage distribution system are required that it recover 0.5-1.0sec in Japanese Grid-interconnection Code. Whether the PCS operates depends on the interconnection point voltage. Therefore, it is important to make the voltage as high as possible under instantaneous voltage drop to prevent PCS shutdowns and blackout.

In Japanese distribution system, there is the voltage management equipment such as load ratio control transformer (LRT) and step voltage regulator (SVR). Recently, a flexible AC transmission system (FACTS) such as static var compensator (SVC), thyristor type step voltage regulators (TVR) appears with the progress of power electronics. We research about the FACTS equipment used silicon carbide (SiC) devices as national (NEDO) project. In this project, those equipment are examined about voltage ride through assist that function is raising voltage under instantaneous voltage drop. Those equipment are categorized into serial type and parallel type and the method of voltage control is different.

Hence, in this paper, we examine the difference of voltage ride through assist performance between serial type and parallel type by experimental result at testing field. This field is same as Japanese distribution system and it can cause various power system fault.

Virtual Synchronous Generator Control and Grid Voltage Control by Reactive Power Coordinated for PV Plant Inverter
Submission-ID 141
Masaya Tsuyuki, Yutaka Ota, Tatsuhito Nakajima
Tokyo City University, Japan
In recent years, the amount of distributed energy resources (DER) introduced in power grids has been expanding, for environmental preservation purposes. Grid connected inverters for DER, however, do not have rotating machine characteristics, unlike conventionally-used synchronous generators. There is serious concern, therefore, that power system stability may decrease at the time of disturbance occurrence. As a solution to this stability problem, virtual synchronous generator (VSG) control has been widely proposed. This paper proposes a new VSG control method for a photovoltaic (PV) plant inverter with a battery storage provided. The originality is that VSG control and governor control for active power and grid voltage control (Automatic Voltage Regulator: AVR) for reactive power are applied together and operated coordinately. PSCAD simulation case studies were carried out intensively, and satisfactory results were obtained.

The proposed VSG control method is based on the swing equation of synchronous generator expressing dynamic characteristics of the generator. The damping and the synchronizing terms are taken into consideration in equation (1). The equation is modeled by transfer functions in a VSG control block.

(M/ωn)(d2δ/dt2)+(D/ωn)(dδ/dt)+Kδ=Pin−Pout (1)

The virtual rotor angle δ is calculated by equation (1), and consecutively, the inverter output power command Pref is calculated by the rotor angle. VSG control and governor control are used as active power control methods, and system voltage control (AVR) is used as reactive power control method.

Simulation case studies using PSCAD/EMTDC were carried out to validate the proposed control method. A detailed model was originally made, including circuits and control systems of a virtual synchronous generator inverter, a synchronous generator and its exciter system. In the analysis, on the other hand, it was assumed that a dc voltage source was connected to the inverter dc side. The simulation condition to examine inverter and generator behaviors was that in a parallel operation of an inverter and a synchronous generator connected to a one-machine and infinite-bus system, a three line-to-ground fault occurred in one circuit of a double-circuit transmission line, and the fault was cleared by breakers. The following four cases were selected to verify the difference in power system stability; (i) VSG and AVR, (ii) APR (constant P) and AVR, (iii) VSG and AQR (constant Q), and (iv) APR and AQR, by checking synchronous generator and inverter dynamic behaviors.

From the simulation results, it was observed that power system stability is highest in case (i), followed by case (ii), case (iii), and lowest in case (iv). It was concluded that, although either VSG control alone or AVR control alone shows a certain system stabilizing effect, the stabilizing effect is further improved by using VSG and AVR in combination.

Approach to determine the effect of local flexibility options within the framework of a smart market platform
Submission-ID 159
Thomas Estermann, Mathias Müller, Simon Köppl
Forschungsstelle für Energiewirtschaft e.V., Germany
The progressive transformation of the generation and consumption landscape is more and more resulting in a complex system. Due to the expansion of decentralized renewable energy generation and an increasing penetration of electric vehicles and heat generators, an increasing load on the distribution networks is taken place. These developments can lead to congestions in distribution grids especially due to high simultaneity.

One possible solution to face this challenge is to use decentral flexibility. There is the possibility that several options are effective for the solution of a local grid congestion. The process to evaluate, which flexibility option should be used, is part of a so-called smart market platform. However, to cover the requirements of the existing market roles, the platform should not have all the necessary data (e.g. grid data) to perform a sufficient load flow calculation, which would be necessary for the technical selection process. These data are only available at the network operator and should be not a requirement to operate the smart market platform.

This paper describes a methodical approach as an alternative to load flow calculations to determine the impact of local flexibility to solve grid bottlenecks while taking the grid restrictions into account. The approach contains so-called effectivity matrices which indicate the electrical effect of a power change (of a potential flexibility option) in advance at a special point in the grid on all other points (potential grid congestion) of this grid area. The platform intersects the effectivity matrices with the available flexibility options, which allows the technical effect of a flexibility option on a grid congestions to be determined approximately without load flow calculation. In addition, the network operator creates these matrices once and does not have to perform load flow calculations for every grid congestion and every use of decentralized flexibility.

Different network bottlenecks require different effectivity matrices, which also differ in their complexity and their use case. If the use case of the platform is an overload of the current, the matrix only determines whether there is an electrical connection between the flexibility option and the bottleneck (‘1’ in the matrix) or not (‘0’ in the matrix). For meshed networks or to solve voltage band problems, the calculation of the matrices is more complex. In this case, the distance between local congestion and flexibility option, the operating state of the grid as well as the grid impedance must be taken into account. To be able to classify the determined results by using these effectivity matrices, the accuracy is compared with the results of load flow calculations.

The presented approach is one of the basic functions for the technical selection of different flexibility options to solve local grid congestions within the framework of a smart market platform without detailed knowledge of the network topology.

Enhanced feed-in management in low and medium voltage distribution grids for PV integration and ancillary service provision – Experience from a field test
Submission-ID 162
Linda Rupp 1, Katharina Volk 1, Joe Imfeld 2, Martin Konermann 1
1 Netze BW GmbH, Germany
2 Landis+Gyr AG, Switzerland
To manage the increasing volatile feed-in due to renewable energy resources while operating the distribution grids within their physical limits is one of the main challenges facing distribution system operators. In addition, due to the decreasing number of conventional power plants the importance of generation at the distribution level for overall system stability grows. This paper presents automated capacity management, enhanced power flow and voltage control methods that are implemented in a decentralized automation system, the Regional Energy Management System. It monitors the medium and low voltage grids which includes the collection and evaluation of measured data to initiate control actions automatically. Based on the results of a field test in a rural medium voltage grid section, it can be demonstrated that grid congestion can be reliably detected, and PV systems are curtailed as needed. Furthermore, enhanced methods to provide a predefined power flow at a reference point are tested and evaluated. These methods can be used to provide aggregated feed-in management to the upstream grid section or control decentralized generation as a wide area power plant.

Evaluation of Intra-Day Distributed PV Forecasts with Customized Metrics Based on Operational Grid Management Decision-making Scenarios
Submission-ID 165
John Zack
UL AWS Truepower, United States
Short-term generation forecasting is recognized as a valuable tool for managing the impact of the variability of solar and wind generation on electric grid systems. However, in addition to accuracy, the value of generation forecasts in an operational environment depends on a number of factors such as the shape and variability of the demand profile, the operating characteristics and cost of available non-renewable generation resources, and constraints imposed by the operating rules associated with individual generation assets. This causes the value of forecasts to the decision-making process to vary significantly within a day and between days. Thus, forecast performance metrics that measure the typical error, such as the Root Mean Square Error (RMSE), often are not a good indicator of the value of the forecasts to the decision-making process and the reduction of integration costs.

In order to address this issue in a specific setting, a study was conducted on an island grid system to examine the types of decisions in which intra-day forecast information was used and the aspects of forecast performance that were critical to positive outcomes for those decisions. The venue for this analysis was the system operated by the Hawaii Electric Light Company (HELCO) on the Big Island of Hawaii. The system demand typically ranges from 90 MW to 180 MW. The system includes 31 MW of wind generation capacity, about 90 MW of distributed photovoltaic (PV) capacity and 16 MW of run-of-river hydro generation capability.

Three critical decision-making scenarios and time frames that are impacted by the variability of renewable generation were identified. These are (1) pre-sunrise preparation for the morning net load peak and the subsequent PV-induced decrease of net load to the mid-day minimum, (2) mid-morning positioning of the system to handle potential large ramp sin the net load caused by ramps in the distributed PV production, and (3) early afternoon preparation for he evening peak. The attributes of solar and wind forecast performance that are critical to each scenario are different.

A customized category-based, event-oriented forecast evaluation schemes was developed for each decision-making scenario. The objective was to provide more relevant forecast performance information to users and also incentive and guidance to forecast providers to focus on improving the most operationally relevant aspects of forecast performance.

The presentation will include: (1) an overview of the generation mix and load profiles on the HELCO system, (2) a summary of the key daily operational decision-making scenarios and time frames with specific case examples of decisions, forecast inputs and outcomes, (3) a description of the customized category-based forecast evaluation scheme for each scenario and (4) a comparison of forecast performance based on the customized evaluation scheme and a traditional set of performance metrics such as RMSE.

Pre-Certification of Grid Code Compliance for Solar Inverters with an Automated Controller-Hardware-in-the-Loop Test Environment
Submission-ID 169
Roland Bruendlinger 1, Johannes Stöckl 1, Zoran Miletic 1, Ron Ablinger 1, Fabian Leimgruber 1, Jay Johnson 2, Jane Shi 3
1 AIT Austrian Institute of Technology, Vienna, Austria
2 Sandia National Laboratories, Albuquerque, NM, United States
3 EPRI, Knoxville, TN, United States
Following the massive increase of the share of distributed energy resources (DER), it has become common consensus that DER need to be fully integrated into system operation to maintain stability and reliability. Accordingly, recent revisions of European as well as International Grid Codes require single generators as well as complete plants to provide a broad range of functions to support the local distribution grid as well as the overall power system. These requirements are complemented with specifications for the robustness of the generators against disturbances, as well as for communication and controllability.

The complexity of the requirements, rapid changes and tight application deadlines set in the documents have led to increased pressure on manufacturers of generating equipment such as PV converters. To qualify for the application, interoperability and performance of the product has to be validated to a wide range of different regional requirements, which are often set on a case-by-case during the planning process.

Traditional lab based testing methods defined in current certification procedures are challenged by the increasing number and combinations of individual tests, limited availability of power equipment in the laboratories and most importantly time and cost constraints.

To overcome these limitations, a test platform has been developed which implements the full set of test items defined in widely used specifications such as VDE 0124-100, FGW TR3 or UL1741SA using a controller hardware-in-the-loop (C-HIL) approach. The platform consists of a software automating the test procedures ("Pre-Cert Toolbox"), converter controller hardware connected to a digital real-time simulation system (DRTS) and a simulation model of the power stage, grid and DC sources. Using the Pre-Cert Toolbox, grid code compliance as well as interoperability and controllability can be validated. Test items are implemented, executed analysed and reported in the exact way as defined in the specifications.

The paper describes the implementation of the C-HIL test platform and highlights its application during design, prototyping and certification. The results obtained from C-HIL tests are successfully validated using setups both connected to a DRTS (Typhoon HIL) running a model of the converter power stage. Tests include

  • Active and reactive power control Q(U), P(f)…
  • Voltage and frequency ride-through
  • Interface Protection, including Anti-Islanding
  • Communication capabilities as per SunSpec

The results show that the C-HIL approach is able to provide significant benefits compared traditional full-scale laboratory testing, allows faster design iterations and reduces the need for cost-intensive power equipment. In addition, complex communication and interoperability requirements can be evaluated individually based on site-specific characteristics prior to the installation of the generator.

Transmission, Distribution and Markets: Coordination needs in the European Power System
Submission-ID 170
Eamonn Lannoye 1, Alison O'Connell 1, Jens Boemer 2, Michael Doering 3
1 EPRI International, Ireland
2 EPRI, United States
3 EcoFys GmBH, Germany
Policy in Europe, at the regional, national and municipal level, has laid out the vision and outline for the energy transition in Europe. A central tenet of that vision is the empowerment of customers and the shift from centralised to localised generation as has been well documented. The challenge emerging from this transition for system operators is to determine how best to evolve their practices to ensure that customers and producers have access to a reliable, high quality, safe, affordable and sustainable electric power system.

This paper focuses on the functions carried out at present by system operators of both transmission and distribution and highlights areas where those decision support analyses and control actions must be adapted to take account of the significant growth of distributed energy resource (DER). By understanding what the data requirements are needed by each actor, the paper then looks at processes for coordination between transmission and distribution system operators and balance responsible parties (BRPs) to ensure that the appropriate data is available when required by each actor. The paper highlights some recent initiatives and examples of demonstration or pilot projects which will pave the way for this cooperation to continue.

The paper also seeks to highlight when the tipping point arises when new types of information are required by each actor to enable the secure functioning of the system. As investment decisions in data management made now will impact the effectiveness of the coordination later down the line, the paper also seeks to highlight the relative merits of different approaches to data management and cooperation and concludes with a set of key priorities for consideration when building the relationships between TSOs, DSOs and BRPs.

Self-consumption and self-sufficiency level in an energy cell with a high penetration of PV and a hybrid battery storage system
Submission-ID 171
Lukas Held 1, Michael Armbruster 2, Martin Zimmerlin 1, Gilles Weber 1, Michael R. Suriyah 1, Thomas Leibfried 1, Rüdiger Höche 2
1 Karlsruhe Institute of Technology, Germany
2 Stadtwerke Bühl GmbH, Germany
In this paper, results of the demonstration project Hybrid-Optimal are presented. The project partners are installing a hybrid battery storage, which consists of a 5kW/ 45kWh Vanadium Redox Flow Battery and a 40kW/ 56kWh Lithium-Ion Battery, in an energy cell with high photovoltaic (PV) generation. In this publication, the self-consumption and self-sufficiency of the energy cell with and without the hybrid battery storage system (HBSS) is calculated. Additionally necessary self-consumption levels for profitable operation are determined.

System imbalance from solar energy trading
Submission-ID 185
Tomas Landelius 1, Sandra Andersson 1, Roger Abrahamsson 2
1 Swedish Meteorological and Hydrological Institute (SMHI), Sweden
2 Tekniska Verken Linköping Nät AB, Sweden
The aim of this paper is to study how the use of an advanced trading method, like the optimal quantile strategy, may effect the balance between generation and consumption at the power system level when trading solar PV power on the Nordic Power Exchange. In order to do this we first developed a set of PV power forecast models. Numerical weather prediction data together with power measurements at 210 PV installations, in the regional network operated by Tekniska Verken Link ̈oping N ̈at AB, were used for estimation and evaluation. Linear and non-linear regression, the latter in terms of an artificial neural network, both resulted in an RMSE, normalized with installed power, of about 6 %. Second we used the neural network to perform a three month simulation experiment on the Nord Pool Elspot day-ahead market. Strategies based on deterministic forecasts were compared with the use of the optimal quantile, based on ensemble forecasts of the power probability distribution. The optimal quantile strategy resulted in an increased revenue of around 2 % but also in an increased imbalance between contracted and produced energy of almost 20 %. The imbalanced part of the power production for the optimal quantile strategy was about one third. A similar study, on trading wind power with the same strategy from a hypothetical plant on the Nord pool market, showed that about half of the traded energy became imbalanced.

Simplification Methods for Optimal Dimensioning of Energy Storage Systems and Heating Devices using Time Series Load and Infeed Data
Submission-ID 192
Martin Zimmerlin, Marek Fritz, Lukas Held, Michael Suriyah, Thomas Leibfried
Karlsruhe Institute of Technology (KIT), Germany
In this paper simplification methods for optimal planning of home energy systems are presented. The planning process is performed based on time series data of solar infeed, power demand and heating demand. The resulting linear optimization problems aims at minimizing the overall cost considering investment and operating cost/revenue. The results show that the presented simplification method of characteristic weeks is able to reduce computational burden without distorting the results too much.

Evaluation of datasets and methods to derive 3D building models and their influence on PV power integration studies
Submission-ID 193
David Lingfors, Joakim Widén
Built Environment Energy Systems Group, Dept. of Engineering sciences, Uppsala University, Sweden
A Geographic Information System (GIS) is a powerful tool for studying the impact of building-applied PV systems on the power system. The grid operator can be advised on exactly which parts of the grid may suffer from problems with operational performance (e.g., over-voltages and overcurrents). However, this requires the PV power generation to be realistically modelled, for which a crucial first step is to find representative building models. In this study, we compare the accuracy of two types of aerial data, LiDAR (Light Detection And Ranging) and photogrammetry data, and two methods for identifying building models from these data, a rasterbased and a vector-based model. The results show that with photogrammetry data, the roof topology of buildings is not identified correctly as often as with LiDAR data, and the vector-based method gives a far better representation of the roofs than the raster-based method. We exemplify this by comparing the results of power-flow simulations on a distribution grid with about 5000 customers where PV systems have been deployed on the roofs according to the two different methods. For the raster-based method the PV power potential is almost four times higher than for the vector-based method, which overestimates impacts on the simulated performance of the grid. The conclusion of the study is therefore that for accurate simulations of the impact of building-applied PV on grid performance based on GIS data, the proposed vector-based method should be used, rather than the raster-based, and it should be based on LiDAR data rather than photogrammetry data.

The role of climate variability on the assessment of roadmaps for power systems with high renewable penetration
Submission-ID 194
Raquel Figueiredo, Pedro Nunes, Miguel C. Brito
Instituto Dom Luiz, Faculdade de Ciências, Universidade de Lisboa, 1749-016 Lisboa, Portugal, Portugal
The spreading concern about the consequences of carbon-intensive activities across all socio‑economic sectors is accelerating the path towards emission-free power systems. This has led to the establishment of targets towards the massive deployment of renewable-based systems. To meet those goals, many countries and communities have been investing heavily in renewables and storage technologies. However, the long-term impact of such measures is uncertain and difficult to assess, since the performance of renewables is strongly dependent on the future climate.

Anthropogenic climate change is well documented in the literature; a common and well-accepted feature of future climate is that it will be further uncertain and variable. Many climate models try to characterize the climate of the future, but as they depart from different assumptions and modelling approaches of atmospheric processes, each model has its own trace for the future.

The proposed work uses the Portuguese power system in 2050 as a case study to assess the impact of the uncertain climate on an electric system with high penetration of renewable energy. We use an hourly-based power system modelling tool (EnergyPLAN) to simulate the power system. Regarding climate, we use two emission scenarios (RCP4.5 and RCP8.5) and the CORDEX climate models – with a one-day and three-hourly resolution interpolated to hourly time-series. For each climate model and each emission scenario, the Portuguese power system is simulated eleven times, accounting for the period 2045-2055. During these simulations, all the system characteristics (installed capacities, load demand, electric mobility assumptions, etc.) are kept constant. The generation of each supply source, the renewables’ electricity share and the CO2 emissions of the system are determined along with the needs cross-border interconnection capacity.

The results allow to quantify how the system operation changes with climate variability. As it was expected, since the modelled power system has a high penetration of variable renewables, the implications are large: following the current renewable energy roadmaps, the Portuguese power system will have in 2050 a renewable electricity share ranging between 78.4 and 95.3% (correspondent CO2 emissions are 1.5 and 8.6 Mton/year), which compares with the 90.9% that the system would have if the 2050 climate was as in the present. Thus, considering climate variability, the renewable electricity share could decrease by 12 percentage points and the CO2 emissions could double.

These results show that future climate uncertainty and interannual variability may jeopardize attaining CO2 emission targets, highlighting the importance of introducing further resilience to renewable-based power systems to achieve a truly sustainable energy system.

Performance Simulation of Integrated Solar Combined Cycle Power Plant which Used Direct Steam Generator Method
Submission-ID 203
Maher Salamah
Ministry of electricity, Iraq-Baghdad, Iraq

The huge interest in the exploit of renewable energies and more particularly in solar energy opens interesting prospects for advanced concepts to integrate solar power in a conventional energy generation technologies. Concentrated solar power (CSP) represents a reliable and sustainable source of energy with different outputs that can be used such as integration with conventional combined power plants to enhance the performance and increase the power generation. Integrated Solar Combined Cycle Power Plants (ISCCs), composed of a Concentrated Solar Power (CSP) plant and a Natural Gas Combined Cycle (NGCC) power plant, have been recently introduced in the power generation sector as a technology with the potential to a very efficient use of the solar and fossil resources .Recently, the direct steam generator method has been introduced as a cheaper way for integrate combined power plants with solar energy field. In the present work, development combined cycle power plant in north of Iraq (225MW) to integrated solar combined power plant is proposed and the preferred method for converting solar energy into electricity is applied by using solar parabolic trough collectors. The mathematical model used in this work is based on governing equations and simulation under various intensity solar irradiations at day hours. Electricity production, Thermal efficiency, net solar electricity, specific fuel consumption and CO2 emissions are the main parameters to evaluate the performance of each component of the solar power station through four months (March, June, September and December) in the year. The simulation results shown that the hybrid power plant can provide about 225 MW with efficiency equals 55% at night. At daytime, the net solar electricity can be very interested and reach the limit of 27 MW and the efficiency of the ISCCS will be above the combined cycle regime by 12%, which are correspond to 252MW and 61.5%, respectively at a highest value of Direct normal irradiation through the summer. Reducing specific fuel consumption and CO2 emissions was about 10% for each one comparison to conventional combined power plant values at night.

An integrated Energy Management System for improved Solar Integration and Storage in Cities
Submission-ID 210
Ulrich Leopold, Gerald Arnould
Luxembourg Institute of Science and Technology, Luxembourg

Across North West Europe, cities are increasingly investing in solar energy production and charging infrastructure for electric vehicles. However, the management systems for energy generation, energy utilisation, energy storage and electric vehicle charging are currently separated from each other. In addition, electric vehicles are mostly powered by fossil fuel generated electricity. At the same time, solar energy is inefficiently utilised because production and demand are not synchronised across the city. The results are high costs and CO2 emissions due to energy inefficiencies.


Within the EU project CleanMobileEnergy, partners are implementing an intelligent and interoperable Energy Management System (iEMS) to balance both demand and fluctuating supply of renewable energies in cities. By interconnecting energy production and distribution systems with temporary local storage solutions, the tool will reduce the current gap between energy production and consumption peaks.

Driven by the partner cities, a strong focus is on the development of the energy management system itself. The iEMS will be based on the iGuess® software technologies developed by LIST, open standards and interfaces and an existing operational solution. Based on geospatial software technologies, interoperability and scalability, renewable energy and eMobility are integrated through the design and development of an open system architecture and its specifications. An operational core system is procured to which researchers will add intelligent and interoperable tools and interfaces to use such system effectively across different cities in Europe.

One generic transnational iEMS will be adapted to 4 specific city pilots in Arnhem (NL), London (UK), Schwäbisch Gmünd (DE) and Nottingham (UK). These pilots range from small towns to large cities. The 4 city pilots cover solar energy amongst other types of renewable energy, storage and electric vehicles as well as different contexts and diverse city environments.


CleanMobilEnergy will make it possible for solar energy sources to be used locally, so electric vehicles can be charged with renewable energy offered at an optimum price. Electrical energy from the grid will only be required when prices are low or renewable energy sources are not available. The intelligent Energy Management System monitors and help to optimise demand, supply and storage.

One generic transnational interoperable Energy Management System solution will be developed that, in the long term, could be implemented across the EU, managing districts, small cities, or major metropolitan areas. European cities will therefore be able to better manage their energy in an effective manner, allowing them to further uptake of renewable energy, limit their fossil fuel requirements and reduce their greenhouse gas emissions.

Wind Power Generation power plant for the CSIR Pretoria Campus
Submission-ID 213
Sengiphile Simelane
Programme Manager, South Africa
The Council for Scientific and Industrial Research (CSIRE) in South Africa aims to create an independently energy supplied site at the main campus by providing energy from the three primary energy sources namely: solar, wind and biogas. This programme will stand as a real-world research platform for designing and operating a primarily renewables-based energy system at the lowest possible cost in R/kWh. Significantly, the annual load requirement for the CSIR main campus in Pretoria is 30 GWh.

As per the wind supply projected requirement of 2-3 MW, the EAC management team commissioned a study to ascertain whether the available wind resource is suitable for a wind turbine generator (WTG) to be installed. Furthermore, the purpose of this study is to determine the feasibility of installing a WTG at one of the available alternative CSIR sites across the country.

The study considers possible options for installing a WTG, three different turbines on different sites with varying average wind speeds. The CSIR Pretoria main campus sites have been modelled in windPRO © with the alternative sites being described using high level calculation results.

The study considers a technical viewpoint, and evaluates annual energy production (AEP). It also looks into the capacity factor. Some sites are remotely viable such as the Port Elizabeth (PE) having an AEP of 6244 MWh/a resulting in a capacity factor of 33 %. Environmental impact assessments is performed for all sites and significant challenges are foreseen in this regard from a project timeline perspective as well. Aesthetically and permitting and zoning is also considered as public and bureaucratic challenges are imminent.

An economic viewpoint, capital investment costs (CIC) for all options is considered within a defined range. The most significant inclusion in the cost / benefit analysis are the tangible and intangible factors of using the turbine as a showcase and research tool. The CSIR Pretoria site performs best in this regard.

Solar energy for residential electric vehicle charging in Northern Norway – a feasibility study
Submission-ID 214
Clara Good, Tobias Boström
UiT The Arctic University of Norway, Norway
This paper presents a study of the potential for using photovoltaic (PV) solar energy systems for residential charging of electric vehicles (EVs) in Northern Norway. The objective is to investigate the load match between PV yield and uncontrolled EV charging, in terms of self-consumption and self-sufficiency. The load profile for EV charging is retrieved from a study by the Norwegian Water Resources and Energy Directorate (NVE), based on measurements and a survey sent to EV owners. An adjusted example EV profile that better represents a single household is also proposed. Other household loads are taken into account using measured data from ten single-family buildings in Tromsø, retrieved from local power company Troms Kraft. The PV yield is simulated for roof-mounted and façade-mounted 4.2 kWp system with different orientations, using PVsyst. The results show that the load match between PV yield and uncontrolled EV charging is poor, as PV power has a peak at noon and the EV charging is highest during afternoon and night-time. A design option for increased load-match (but lower total yield) is mount the PV system facing west, since the PV power peak is shifted towards the afternoon. Solutions for increasing the load match, provide autonomy and reduce negative impacts on the grid are discussed, for example the use of residential battery storage and controlled EV charging. Based on the results, the authors propose that more focus is given to workplace charging combined with solar energy, since this would increase the load match significantly.

Impacts of Distributed Solar Advanced Inverters on Transmission Voltage and Reactive Power
Submission-ID 227
Bryan Palmintier, Ibrahim Krad
National Renewable Energy Laboratory (NREL), United States
This paper uses large-scale, integrated transmission-distribution (T-D) co-simulation (hundreds of full-scale feeders combined with complete transmission) to capture the impacts of advanced inverters on reactive power and resulting voltage at the TSO-DSO boundary for US-style electric systems. This is an area of growing concern for very high penetration distributed generation from solar photovoltaics (DGPV), which increasingly use advanced controls with various power factor and volt-VAR strategies to manage local voltage on the distribution system. However, this local voltage support can also potentially increase reactive power demand from the bulk system impacting power flow, losses, and transmission level voltages.

Extensive past research on advanced inverters providing volt-VAR control has focused on distribution-level benefits (e.g. [1]) while studies of transmission voltage impacts have been largely limited to the control of central, utility scale PV plants (e.g. [2]). Past integrated T-D work, including [3] by the author, hints at the ability for distributed PV to impact transmission voltages, but does not explore the trade-offs among various options for inverters, such as a range of fixed power factors and alternate volt-VAR profiles.

In this paper, we utilize the high-performance IGMS-HELICS co-simulation infrastructure [4], [5] to overcome this gap. We capture the AC powerflow interactions between T&D with varying penetrations and alternate advanced inverter modes to explore their impacts on transmission reactive power and voltages. The IGMS-HELICS framework couples transmission powerflow in MATPOWER with hundreds of distribution feeder powerflows in GridLAB-D to uniquely examine cases where high-pen DGPV transitions from “voltage-taker” to“voltage-maker,” such that the impacts of DGPV on transmission voltage create a feedback loop that in turn impacts the distribution voltage and hence DGPV reactive power.


[1] F. Ding et al., “Voltage support study of smart PV inverters on a high-photovoltaic penetration utility distribution feeder,” in Photovoltaic Specialists Conference (PVSC), 2016, pp. 1375–1380.

[2] A. Ellis et al., “Review of existing reactive power requirements for variable generation,” in 2012 IEEE PES General Meeting, 2012, pp. 1–7.

[3] B. Palmintier et al., “On the Path to SunShot: Emerging Issues and Challenges in Integrating Solar with the Distribution System,” National Renewable Energy Laboratory, Golden, CO, NREL/TP-5D00-65331, May 2016.

[4] B. Palmintier et al., “IGMS: An Integrated ISO-to-Appliance Scale Grid Modeling System,” IEEE Trans. Smart Grid, vol. 8, no. 3, pp. 1525–1534, Sep. 2016.

[5] B. Palmintier, D. Krishnamurthy, P. Top, S. Smith, J. Daily, and J. Fuller, “Design of the HELICS High-Performance Transmission-Distribution-Communication-Market Co-Simulation Framework,” Workshop on Modeling and Simulation of Cyber-Physical Energy Systems, Pittsburgh, PA, 2017.

Coordinated Reactive Power Control of Solar PV for Thermal and Voltage Constraint Management in Distribution Network
Submission-ID 233
Hieu Nguyen 1, 2, Takao Tsuji 1
1 Yokohama National University, Japan
2 Vietnam National University of Agriculture, Japan
The rapid proliferation of solar photovoltaic (PV) units poses operational challenges for the low and medium voltage network assets along with voltage limit violations at the connection points. To tackle these issues, inverter-based methods are preferred to be used by active and/ or reactive power control. However, curtailing active power reduces customers' revenue while large reactive power absorption required from the upper network might overload the transformer, especially if the same method is applied in all low voltage (LV) network. Additionally, it requires low power factor (PF) operation in the upper network. In this paper, PVs located near the substation will inject reactive power to compensate reactive power absorbed by PVs staying further in the network. By this expected cooperation, not only overvoltage mitigation but also minimization of reactive power consumed from the upper network can be achieved.

Keywords-component: photovoltaic (PV) system; reactive power control; voltage control; thermal management;

Calculating the Hosting Capacity of Electrical Network with high penetration of Solar PV
Submission-ID 234
Nicholas Etherden 1, David Lingfors 2, Kjetil Kvamme 3
1 Vattenfall R&D, Sweden
2 Uppsala University, Sweden
3 Powel A/S, Norway
The Hosting Capacity (HC) is defined as the maximum amount of new production or consumption that can be connected without endangering the reliability or quality for other customers. This paper presents a method for estimating the grid HC of individual nodes in medium voltage (MV) and low voltage (LV) networks with various amounts of PV installed.

Solar radiation on roof surfaces for a set of buildings in the studied grid were calculated where 3D representation of buildings are created. This allows hourly radiation time-series to be produced for each PV module with potential production over 950 kWh/m2, yr. Time series are associated with meters within the property boundary and loaded to a Network Information System (NIS), with integrated electrical analysis and calculation capabilities.

The solar model returns times-series of PV power generation, which are influenced by both local weather conditions and shading. The time series are used in a commercial NIS tool. Results from a real utility network will be presented. The method calculates the HC of each MV and LV node in a selected grid section and lists the grid components limiting the HC. The net power flow with the overlying MV network is derived for different proportion of roofs with PV installed. Simultaneously the increased of the HC is calculated in NIS for some of the most common DSO mitigation actions, including . This enables accurate estimates of the network investments required to support different penetration levels of PV within a network.

Next, power flow analysis are conducted with a set proportion of the roofs having solar cells. (Roofs are selected according to the return-of-investment, according to the utilities commercial residential PV offering.) In the study, the power flow is calculated in NIS for a calendar year. Voltage at each node can be seen for each hour, as well as voltage profiles along LV circuit under each 10/0.4 kV substation for different amounts of PV. The amount of installed capacity possible without violating component design criteria or regulation such as EN 50160 assigned to each network node as the HC limits.

Calculations were performed in NIS for network with 534 customers and 51 MV substations with 10, 20 and 30% of the roofs with best irradiation having PV modules. It can be seen how a reverse power flow appears mid-day during the summer with 20 % of the roofs having PV (between 3,2 and 43,2 kW per roof).

The work demonstrates not only how that the hosting capacity can be calculated in commercial NIS, but simultaneously also the increased amount of PV acceptable when applying three of the most common mitigation solutions: increasing cable area, change of MV transformer tap position and change of PV power factor. This means that the DSO is not only informed about what amount of solar cells that can be accepted without network enforcements, but also the most effective network investment to allow more solar energy in a local energy community.

Development of Nowcasting Method of Irradiance Distribution based on Kriging and All-sky Image Information
Submission-ID 239
Takeyoshi Kato, Masaki Imanaka, Muneaki Kurimoto, Shigeyuki Sugimoto
Nagoya University, Japan
This paper proposes a novel nowcasting method of irradiance distribution based on combination of the Kriging method as an interpolation technique and the visible information of cloud distribution by an all-sky image capture.

The conventional Kriging method estimates the irradiance at arbitrary points without observation as the weighted average of the irradiance at several observation points. The weight factors are determined based on the relation between the irradiance similarity and the distance of various pairs of observation points. Therefore, the conventional Kriging method is not useful when the cloud distribution suddenly changes independently on the distance, requiring the increased number of observation points for the higher spatial resolution. To overcome the disadvantage of the conventional method, our proposed method adds some virtual observation points and applies the Kriging method using both the observed irradiance and the estimated irradiance based on the color information over 49 virtual observation points are placed in a grid pattern.

Because the effective area of all-sky image changes depending on the height of cloud, it is crucial to take the effective area from the whole captured image, while the estimation of cloud height using single all-sky image is very difficult. In the proposed method, the useful are is chosen by evaluating the similarity of observed irradiance and estimated irradiance at the several observation points. Then, using the estimated irradiance at 49 virtual observation points and the observed irradiance at several points, the proposed method estimates the spatial distribution of irradiance in the target area by using the Kriging method.

This study evaluated the performance of proposed method by using the all-sky image captured at Nagoya University and the observed irradiance at 7 points in 8×8 km area around Nagoya University at the same time of the image capture. In the captured image used on July 13th 2014, a large cloud overcasts the western area, indicating the small irradiance. On the other hand, because some broken clouds cover the rest area, observed irradiance is not so small except one points. Because of no observation points in the western area, the estimated irradiance using the conventional Kriging method is affected by the observed large irradiance at neighboring points, resulting in the large estimation error. On the other hand, by placing several virtual points and applying the observed small irradiance at the east-end point in the rest area covered by small cloud, the small irradiance in the Eastern area is successfully estimated by using the proposed method.

The performance of proposed method has been evaluated by comparing a Kriging-based method using observed only irradiance at several points only. As a result, it is revealed that the proposed method is useful for estimating the irradiance distribution in the situation that some broken-clouds float over the target area.

Stability of grid-connected photovoltaic inverters during and after low voltage ride through
Submission-ID 248
Ziqian Zhang 1, Robert Schürhuber 1, Lothar Fickert 1, Yongming Zhang 2
1 Institute of Electrical Power Systems Graz University of Technology, Austria
2 Shanghai DianJi University, China
In the practical application of PV grid-connected inverters, the failure of low voltage ride through has repeatedly occurred. Even if inverter model passed the grid compliance testing in test laboratories, multiple cases been observed where inverters lost stability during voltage dips in practice. The reason for this behavior were stability problems of the PLL (phase-locked-loop) circuit implemented in the converter control. This circuit is needed to set the phase reference for the inverter control. In this article, a detailed mathematical modeling of the inverter phase-locked loop system is carried out and analyzed. Based on the second order nonlinear differential equation, an intuitive graphical method is used to analyze the system. This analysis concludes that during and after low voltage ride through, stability of inverter depends on interaction of several control loops in the inverter. This interaction can lead to instability after fault clearing, since the phase-locked loop has left its area of stability. Through a power hardware-in-the-loop test system, this article reproduces this phenomenon in laboratory.

Power quality measurement campaign at a Jordan LV grid and determination of the influence of a large PV plant
Submission-ID 250
Daniel Masendorf 1, Eckehard Tröster 1, Stanko Jankovic 2
1 Energynautics GmbH, Germany
2 GOPA - International Energy Consultants GmbH, Germany
Power Quality has been measured at several points in the LV grid of a Jordan refugee camp over a period of more than 1 year. During the measurement campaign a 12.9MWp PV plant was connected closeby to the upstream MV grid. The power quality in the grid, the influence of the PV plant and its method of operation on the power quality and lessons learned during the measurement campaign are described in this paper. The focus of the analyses was put on the voltage magnitude,unbalances, THD and flicker.

Experiences with large Grid Forming Inverters on the Island St. Eustatius, Portability to Public Power Grids
Submission-ID 252
Paul-Robert Stankat, Oliver Schömann, Christian Hardt, Andreas Falk, Thorsten Buelo
SMA Solar Technology AG, Germany
Abstract for: 8th Solar Integration Workshop 16-17 October 2018 Stockholm, Sweden

Title: Experiences with large Grid Forming Inverters on the Island St. Eustatius, Portability to Public Power Grids


Oliver Schoemann, Paul-Robert Stankat, Thorsten Buelo, Christian Hardt, Andreas Falk(all SMA Solar Technology AG)

Date: 2018 August 31st

Island grids on LV-level with grid forming inverters in the power region of some kW up to 150 kW are state of the art while larger systems don´t offer a grid forming functionality with battery inverters up to now.

In 2017 SMA installed a battery supported PV-diesel-hybrid system on the Caribbean island St. Eustatius (belongs to the Netherlands, about 4000 inhabitants), able to run the entire electric energy supply system of this island including MV- and LV-distribution grid in both diesel-off-mode and in diesel parallel operation with an average power demand of about 2 MW. By installing the grid forming inverter technology, it was possible to shift the share of PV-generated electricity from about 25% to almost 50%.

During diesel-off-mode several battery inverters are parallelized via frequency and voltage statics. The overall energy management and multiple control functions on system level are conducted by an overlaid plant controller.

The experiences of such a system are displayed in this presentation. The main focus of the presentation is the transient behavior during faults in power- and communication-networks and the quality of voltage and frequency control. Current controlled PV-inverters and voltage controlled battery- inverters operate in parallel in this application.

The algorithms tested during this implementation enable the possibility to run grid forming battery inverters in parallel to the public grid. Such inverters provide virtual inertia and are necessary to replace grid forming synchronous machines (so called “must-run-units”). The second part of the presentation gives an outlook to the possibility of this technology.

An outlook describes next steps, milestones and goals. Regulatory boundaries and further development and research efforts are pointed out.

International Communications Standard for Meeting Smart Energy Grid Code Requirements: IEC 61850-7-420 for DER
Submission-ID 254
Laurent Guise 1, 2, Frances Cleveland 3, 2
1 Schneider-Electric, France
2 IEC TC57 WG17, Switzerland
3 Xanthus Consulting, United States
This paper describes how utilities can benefit from the most recent developments in the leading smart energy communications standard of IEC 61850, which now provides the data model required to implement observability, operational functionality, and controllability of DER that are interconnected to the grid, thus fully supporting the new regional and national network Grid Codes requirements developed in Europe and North America

Evaluation of heat pumps for balancing grids in combination with solar energy production: A Dutch Case Study
Submission-ID 260
Cihan Gercek, Angele Reinders
Department of Design, Production and Management, University of Twente,, Netherlands
This paper analyzes the use patterns of smart grid pilot in the Netherlands. Heat pumps and other appliances’ usage are inspected in order to evaluate their effectiveness, their ability to reduce peak electricity purchase, their effects on self-sufficiency and solar energy local usage. The analysis is based on monitoring data originating from a residential smart grid pilot, PowerMatching City which occurs in the Netherlands. The study took part in the framework of the research project “Co-Evolution of Smart Energy Products and Services” which has been executed in the ERA-Net Smart Grids Plus program. As result, diverse yearly and weekly values are presented to highlight the consumption and production of these smart household’s properties.

Design innovation methods as tools for overcoming barriers in smart energy product design
Submission-ID 262
Alonzo Sierra, Cihan Gerçek, Angèle Reinders
Department of Design, Production and Management, University of Twente, Netherlands
The study of smart energy products and services (SEPS) from a design perspective may provide new insights for overcoming some of the key barriers their development faces at present. This paper aims to review smart grids research to identify the main barriers for SEPS design and their possible solutions, followed by an analysis of some of the design techniques used to develop new SEPSprototypes. Possible measures for removing the barriers identified include focusing on motivating users beyond economic incentives, presenting feedback in a simpler, more visual way and the responsible collection and management of user data. The analysis on design methods revealed the usefulness that each of them had at different stages of the product design process, although the impact of some methods was harder to determine due to the scope of the evaluated projects.

Coordinated power system services from distributed batteries
Submission-ID 268
Per Nørgaard 1, Michalis Florides 2
1 Technical University of Denmark, Denmark
2 University of Cyprus, Cyprus
The increasing level of variable energy generation in the power system (from wind and solar) requires more power and energy flexibility in the power and energy system to balance the power and energy in the systems. Batteries in the power system is one mean to provide flexibility. Batteries provide no energy service. They solely provide power system services. Their energy capacities are typically too low to provide significant energy system services.

Battery systems can provide active and reactive power in both directions, more or less independently of each other and with fast response. The power to / from a battery system will simultaneously affect the local voltage, the local power balance and the global power balance. A battery system can in addition provide virtual inertia and short circuit power for power system stability and system fault handling, respectively. A battery system therefore necessarily simultaneously provides different and interrelated local and global power system impacts and services. Any activation of the power flow to or from a battery will have an impact on the local voltage, the local power flow and the global power flow. The global impact from a single battery system may be small, but the coordinated activation of many batteries distributed within an area can be significant. The optimal coordinated operation of more batteries distributed in a power system is therefore extremely complex.

The design of the power markets defines the business cases for the battery systems and thereby the way the batteries are used and operated. If the power markets and tariffs are not properly designed, we will see sub-optimised solutions, that are not necessarily the optimal system solutions – like the PV + battery ‘behind the meter’ solutions, optimised to reduce the electricity bill.

The paper presents a methodology for a coordinated operation of many batteries distributed mainly in the power distribution system seen from a system perspective (not from a private business perspective), taking into account the various and varying power system service needs – varying in both time and space.

Solar inverter interactions with DC side – some regulatory challenges
Submission-ID 270
Scott Hagaman 1, Jennifer Crisp 1, Hieu Nguyen 2, Tim George 1, Ravidutt Sharma 1
1 DIgSILENT Pacific Pty Ltd, Australia
2 Leeson Group Pty Ltd, Australia
The DC voltage on the photovoltaic (PV) array connected to an inverter plays an important role in the operation of the PV inverter system. The DC voltage is controlled by the maximum power point tracking (MPPT) controller to deliver active power based on the PV array V-P curve. This varies with solar irradiance and PV cell temperature. The DC voltage needs to be kept higher than the peak AC voltage at the inverter terminal for the inverter to operate correctly. It is also used to regulate the PV array output. The DC voltage is thus a function of both the PV array design (solar irradiance and cell temperature) and the inverter side (peak AC voltage).

The need for a margin between the DC voltage and the peak inverter terminal voltage creates situations that may limit the active power and deliver unexpected responses. These can have regulatory compliance issues because, in Australia, the technical standards require ‘continuous uninterrupted operation’ (CUO). In brief, CUO has been interpreted to mean that the active power remains unchanged following power system disturbances, and that reactive power is only changed as required for voltage control for connection point voltage variations within an operating band of 90%-110%. In practice this means that additional inverters are required to meet CUO and the ratio of inverter capacity to rated active power typically exceeds 1.2. In normal operating conditions, the power plant controller must limit inverter operation to a value below the MPPT optimum at high irradiance levels, reducing both active power and yield to a level that allows active power and reactive power capability to be maintained at 90% nominal connection point voltage, without tap changer operation.

For some irradiance/ high temperature conditions, active power may reduce if the inverter ac voltage increases to within the acceptable margin of the DC voltage. Under these conditions, the DC voltage is increased to maintain the required margin between AC peak and DC voltage, consequently curtailing active power. Depending on the AC design of the generating system, the collector voltage can be significantly affected by system voltage variations. Under some feasible conditions, a decrease in inverter voltage caused by a change in the system conditions, can result in an increase in active power output when the active power is initially below maximum available level. This is not usually a power system security issue but in Australia it creates compliance issue, which has the potential to exclude otherwise high performing inverter systems or have high compliance costs.

Very Short-Term Solar Power forecasting Using Ground-based Sky Images
Submission-ID 271
Lasanthika Harshani Dissawa Dissawe Mudiyanselage 1, 2, Ashish Agalgaonkar 1, Duane Robinson 1, Sarath Perera 1, Roshan Godaliyadda 2, Parakrama Ekanayake 2, Janaka Ekanayake 2, 3
1 School of Electrical, Computer and Telecommunications Engineering, Faculty of Engineering and Information Sciences, University of Wollongong, Australia
2 Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Peradeniya, Sri Lanka
3 Cardiff University, United Kingdom
This paper proposes a very-short-term PV power forecasting technique centred on cloud motion estimation employing ground-based sky images. Sequences of sky images taken from a wide-angle camera are processed to identify individual clouds separately and motion vectors of each cloud are identified using a cross-correlation based cloud motion tracking algorithm. Based on the cloud motion vectors, the cloud shadow movement is estimated. The drop in irradiance (and the associated PV power output) levels at a site, where the camera is fixed, is found using two methods namely ray tracing method and the direct method, which are based on the sun occlusion time calculation. By utilising the historical PV power/irradiance measurements with associated cloud and shadow movements, the PV power is forecasted for 5 minutes ahead of time.

Smart Meter Design for e-Health monitoring on the Smart Grid System
Submission-ID 272
Amleset Kelati
Royal Institute of Technology (KTH), Sweden
University Of Turku, Finland
In today's and in the future technology trend there is a lot of interest in the development of smart devices that can be compatible with the IoT system. Internet of Things (IoT) based smart grid present a reliable, quality, secure and costeffective power supply automation of electrical power systems processes. Advancement of information acquisition, storage, processing, transmission and sensor technologies have efficient systems for monitoring, evaluating by analysis, and perform efficient control of the electrical consumptions. In this paper, we propose a design of smart meter for monitoring, load profiling and control the smart grid to improve the life quality of elderly people located in the e-health center by measure the appliances usage information as time, duration and. energy consumption. The smart meter data is used to analyze and monitor the elderly people normal or abnormal behavior and the system have better performance and real-time monitoring capability by automatically logging results and helping the patients always monitoring by their caregivers. .

On the Improvement of Day-ahead Forecasts of Solar Irradiation with Simple Ensembles and Training Data Selection in Japan: A Countrywide Assessment
Submission-ID 273
Joao Gari da Silva Fonseca Junior 1, Fumichika Uno 2, Takashi Oozeki 2, Kazuhiko Ogimoto 1
1 The University of Tokyo, Japan
2 National Institute of Advanced Industrial Science and Technology - AIST, Japan
The objectives of this study are to investigate the effect of two techniques to reduce the errors of day-ahead forecasts of solar irradiation done with machine learning, and to present a countrywide assessment of the current accuracy of such forecasts in Japan. The two techniques investigated were training data selection based on a concept of weather similarity, and the use of simple ensembles devised to reduce the influence of inaccuracies of cloudiness variables used as part of the input data. The countrywide assessment of the forecast error was done using 1 year of hourly day-ahead forecasts from April of 2016 to March of 2017 for 41 locations covering the main regions of Japan from North to South. Using a reference method to assess the forecast errors, typical annual RMSE values ranging from 0.0995 kWh/m2 (Abashiri, Hokkaido) to 0.129 kWh/m2 (in Naha, Okinawa). Regarding the error reduction approaches, both were effective; and combined they yielded maximum error reductions of 7.5% of the MAE and 6.5% of the RMSE, comparing with the reference errors. The skill of the forecasts was also improved, in average, by 7.4% Such results indicate the efficacy of both techniques to reduce the forecast error.

High Penetration of Photovoltaic Energy and Supply-Demand Balance in the Western Japan Grid, with Utilizing Interzone Transmission and Demand Response
Submission-ID 274
Asami Takehama 1, Manabu Utagawa 2
1 Ritsumeikan University, Japan
2 The National Institute of Advanced Industrial Science and Technology (AIST), Japan
1. Introduction

The aim of the study is to describe an energy scenario for large-scale integration of variable renewable energy (VRE) sources in the western Japan grid in 2030 through use of interzone transmission, pumped storage hydropower systems (PSHPs) and demand response with heat pump systems (HP) and electric vehicles (EVs). We develop a simplified model of Unit Commitment with Economic Load Dispatching (UC-ELD) of thermal power plants to evaluate impacts of high penetration of photovoltaic (PV) and wind energies on the supply-demand power balance. The amount of power surplus, the required amount of demand response and control reserve activations beyond individual control zones are evaluated. Our model pursues the reduction of coal power operation as much as possible.

2. Methods

(a) Unit Commitment and Economic Load Dispatch model

Using a simplified UC-ELD model, conventional power units were classified into 22 subgroups by considering techno-economical parameters of generation units, such as load following capability in increasing or decreasing output, lower power output limit, and availability of a control reserve for load-frequency control (LFC). The UC-ELD model uses the aggregated capacity of each subgroup power source. The objective function of the model is minimization of hourly generation cost. The UC-ELD model estimates supply-demand balance in May (low demand period) and August (high demand period) in 2030 in the western Japan grids. The UC-ELD optimization calculation considers fuel costs.

(b) High renewable scenario in 2030

The target capacity of the High case in 2030 is 18 GW of PV and 4.7 GW of wind power in the Kyushu zone, and 5 GW of PV and 2.6 GW of wind power in the Shikoku zone. The High case sets the following assumptions: renewable power is transmitted to interzone lines as a priority; no nuclear power is in operation; the operational capacity of coal power is reduced as much as possible; HP and EV in charging mode are activated in daytime to absorb power surplus of PV output. The LFC control reserve is activated beyond the control zone. PSHP is in pumping mode in the daytime, and in generating mode in the evening (PV Pump-Up).

3. Results and Conclusion

The grid balance in the Kyushu zone in May in the High case shows that around 25% of PV output is absorbed by PSHP and around 20% is transmitted to the Chugoku zone as a priority. A large scale of power oversupply of PV output may occur in May in Kyushu, Shikoku and Chugoku zones. The available capacity of the negative control reserve would be short in the High case. The utilizing PV power curtailment for negative control reserve is recommended. A combination of the use of PV Pump-Up operation, interzone transmission and demand response can manage a part of supply surplus from PV output in the High case.

Impact of inverters with virtual synchronous machine control in low voltage grids
Submission-ID 277
Florian Rauscher, Edwin Rebak, Bernd Engel
Technische Universität Braunschweig Institut für Hochspannungstechnik und el. Energieanlagen - elenia (1), Germany
The constant increase of renewable energy systems share is leading to a change of the dominating feed-in technology in the grids. Whereas conventional power plants are dominated by synchronous generators, renewable energy sources are mainly connected to the grid through inverters. This leads to new challenges regarding ancillary services as well as essential physical qualities which would usually be provided by the characteristic of a typical synchronous generator. In inverter-dominated grids, however, these have to be re-created through new inverter technologies in order for the grid to remain stable.

One of the most important features of a synchronous generator is the inherent power reaction to sudden load variations, which is based on its voltage source behavior. This behavior is important within the very first milliseconds to cover load variations. Opposed to this stands the fundamental difference of renewable energy sources typically acting as a current source. In a scenario where a high penetration of inverter-based generation in grids occurs, the thereby reduced number of synchronous generators and hence the absence of their voltage sources behavior and balancing characteristics, can lead to instability of the grid.

A possible solution for the scenario above is to use inverters as voltage sources to regain an inherent power reaction. The field of application is usually a stand-alone supply in island grids. In order to guarantee a safe coupling with the mains it is necessary to have a superior control emulating the elementary behavior of synchronous generators. This control will provide synthetic inertia in order to reduce frequency deviations. Thus instead of using conventional power plants to provide the services of inertia, inverters with virtual synchronous machines controls (VSM) can be used to stabilize the grids through providing inertia.

This paper illustrates the impact of such inverters with VSM-control in typical low voltage grids as well as its positive effects on the power balancing while focusing on short-term simulations in order to show aspects of the dynamic stability, especially the angle stability and frequency stability.

Exploring Indian solar photovoltaic technological innovation system
Submission-ID 281
Akoijam Amitkumar Singh
Jawaharlal Nehru University (JNU) New Delhi, India, India
This paper is an attempt to understand Indian solar photovoltaic technological innovation system. Globally, solar photovoltaic industry becomes one of the leaders of the renewable energy sector with consistently increasing the power generation. The study is motivated by the need to accelerate the supply of renewable energy with a special focus on solar energy with a vision of affordable and reliable clean energy because it addresses the issue of climate change and fostering sustainable development in India. The economy of the country is one of the fastest growing in the world. The consumption in the energy sector has increased rapidly in the country due to the rapidly growing population and growing economy. About 300 million people (nearly 24%) are not connected to the national electrical grid (World Bank Report, 2014). In this study, we adopt technological innovation system (TIS) framework to understand inducement and blocking mechanisms and to see the future directions pertaining to solar PV technological innovation system. The study adopts mixed methods, both qualitative and quantitative. Methodologically, the study relies on both primary and secondary data. The TIS perspective emphasizes systemic interdependencies between these elements, which give rise to various forms of synergies, such as collective assets on which the different actors can draw but which they could not produce if they worked in isolation. It has seven system functions such as entrepreneurial activities, knowledge development, knowledge diffusion, guidance of the search, market formation, resources formation and creation of legitimacy. The strength of this framework lies on the analysis of different structural components of the innovation system influence the development and diffusion of a specific technology. It provides insights to policymakers by identifying systemic weaknesses that hamper the development and diffusion of solar PV technology in the country. We found that entrepreneurial activities function embraces a fairly accumulated technological innovation system that performs a dynamic result. At the same time, the least accumulated in the system is associated with resource mobilisation. Inconsistent regulations, less interaction in networks, disconnected competitive entities, unpredictable behaviour of the government and the lack of funding are the key blocking mechanism of India’s PV industry development.

Keywords: Solar photovoltaic, Technological Innovation System, Renewable Energy, India, Sustainable Development

Effects on greenhouse gas emissions of introducing lithium-ion batteries for stationary power system applications
Submission-ID 283
Simon Davidsson
Department of Space, Earth and Environment, Chalmers University of Technology., Sweden
As the use of lithium-ion (Li-ion) battery storage for stationary power system applications becomes increasingly common, the environmental effects of this development need to be examined. Using batteries to store power generated from photovoltaic (PV) systems affects the total greenhouse gas (GHG) emissions associated with both the manufacturing and use of the battery. Here, the emissions related to the commissioning of a Li-ion battery and the induced charge-discharge losses are estimated using on a simple battery model. When battery storage enables the use of power that would otherwise be curtailed, the introduction of a battery decreases GHG emissions if the total emissions of the combined PV and battery system is lower than the carbon intensity of the replaced electricity. However, if the PV generation would be used even without the battery, the emissions related to the manufacturing of the battery and the induced charge-discharge losses increases the total GHG emissions associated with the system. Although batteries can have important positive effects on power systems, the full system GHG emissions should be considered before Li-ion batteries are deployed in large-scale for stationary applications.

A comparative analysis of PV markets in Brazil and Sweden
Submission-ID 294
Semida Silveira 1, Wadaed Uturbey 2, Hendrigo Batista da Silva 2, Luciana Marques 2
1 KTH Royal Institute of Technology, Sweden
2 UFMG Federal University of Minas Gerais, Brazil
PV markets in Sweden and Brazil are delayed if compared with other countries that have focused on solar power longer. There are many reasons why it has been so, including perception of the potential, focus of national policies and institutional arrangements among others. Despite that, the markets for solar power are now maturing in both countries. Meanwhile, this is happening in different ways. This paper explores the different conditions for development of PV markets in Sweden and Brazil, particularly grid-connected distributed PV for low-voltage consumers. We investigate institutional arrangements and how these are affecting PV market strategies and its development. In Brazil, around 60 distribution utilities own and operate the distribution grid in exclusive concession areas in a regulated electricity market for small consumers. Grid and energy services are provided by the same distribution utility. Low-voltage consumers pay an energy tariff that includes consumption, grid connection, electricity sector charges and taxes. In Sweden, the retail electricity commercialization and grid services are separated. All consumers can choose their energy provider among nearly 120 energy market retailers. The comparisons highlight the different contexts in which solar power is evolving and explores some of the consequences this may have in the context of technology dissemination and power markets transformation.

DER Integration Study for the German State of Hesse - Methodology and Results for the Medium- and Low-Voltage Level
Submission-ID 297
Alexander Scheidler 1, Jan Ulffers 1, Johannes Dasenbrock 1, Daniel Horst 1, Carsten Pape 1, Martin Braun 1, 2
1 Fraunhofer IEE Kassel, Germany
2 Universität Kassel Kassel, Germany
We present results from a grid integration study conducted for the geographic region of the German federal state of Hesse. Goal of the study is to determine the expected cost for grid reinforcement and expansion in this region for different Distributed Energy Resource (DER) scenarios, to identify effective measures for an advanced DER integration and the identification of needs for action in the regulatory framework. In this paper we focus on the low- and medium-voltage level and present the applied simulation methodology as well as some key results.

Resource Adequacy at Continental Scale: Methodological insights from joint assessments of interconnected North American power systems
Submission-ID 299
Gord Stephen
National Renewable Energy Laboratory, United States
Probabilistic resource adequacy assessment quantifies the risk of unserved demand in a power system resulting from a supply shortfall or deliverability constraint. System parameters including dispatchable generator and transmission line outage rates, variable generation availability, and interregional transfer limits are considered in order to characterize shortfall risk according to metrics such as loss-of-load expectation and expected unserved energy. Such metrics can be applied in turn to quantify the incremental contribution of individual resources towards system-level resource adequacy, most commonly in terms of capacity value.

While resource adequacy assessment has often historically been limited to the smaller footprints of individual grid operators, increasing penetrations of solar and wind power generation and the spatial diversity benefits of pooling such resources across large geographic areas motivate an expanded geographic scope when studying system resource adequacy. Understanding potential impacts from load diversity across time and climate zones in large systems is an additional benefit of performing such analyses across wider geographic extents. The analysis of very-large-scale power systems reveals challenges associated with aspects of traditional resource adequacy assessment methods, including choice of metrics and the quantification of contributions from resources spanning multiple regions. This presentation will outline such issues as well as solutions that have been employed in NREL's work studying large interconnected power systems spanning the North American continent.

Solar ensemble forecasting for distribution power system volt and var control
Submission-ID 310
Bahri Uzunoglu 1, Marianne Blaschek 2, Christian Kusmitsch 2, Alper Terciyanli 3
1 Uppsala University, Division of Electricity, Sweden
2 UBIMET, Austria
3 ENDOKS, Turkey
This paper will present findings of solar ensembles that have been developed for valt and var control for smart distributed smart grids for EU project titled MIDAS. The ensembles address the uncertainty in forecasting. The ensemble simulations are generated by Weather Research and Forecasting method for test sites managed by the project partner Utility compnay. The solar ensembles are presented in the context of control mechanism that can employ ensembles in decision making for voltage and power control for distribution grid. The results of 38 ensembles will be presented for several solar power plant locations. The ensembles are generated as lagged solar ensembles. The findings will be presented and discussed.

Optimal power flow based distributed controller employing ensemble based states
Submission-ID 312
Bahri Uzunoglu
Uppsala University, Division of Electricity, Sweden
This paper will study the power distribution system with solar generation control that develops a control framework to steer the solar output based on suitable linear approximations of the AC optimal power flow problems based on improved ensemble states. The design of the control employs the distributed optimization techniques. The oblective is to manage the flexibility offered by power-electronics-interfaces to manage voltage regulation, power losses, economic benefits to utility and customers. This study develops a power distribution system approach for integration of inverter-interfaced solar plants. The approach develops distributed control algorithm that manages the output powers to solutions of AC optimal power flow problems.

Capabilities of strict power limits at prosumer households for solving network load issues at LV grids: An Austrian Case Study
Submission-ID 313
Stefan Uebermasser, David Reihs, Felix Lehfuss
AIT Austrian Institute of Technology, Austria
Based on the agreements of COP22 and the 2030/2050 strategies of the European Commission, generation of electricity will change significantly during the upcoming decades. As part of this movement, the mass-introduction of distributed energy resources to medium and low voltage power grids already started. Small generation units, such as photovoltaic (PV) systems, but also new technologies such as electric vehicles (EV) and battery electrical stationary storage (BESS) systems are changing the characteristics of the power consumption profile of end-customers. So-called prosumer households (equipped with PV, EV and/or BESS) show a significant different residual load profile than grid connected customers without such systems.

Distribution system capacities, especially at LV grids, are limited and grid connected customers usually have a contractually agreed maximum power value of 2-4 KW per household (depending on the distribution system operator DSO). In Austria the contracted maximum value (over a 15-minute duration) is under “fair use” conditions and not executed strictly. Whilst in the past, only energy and not power was measured at end customer premises, the introduction of automated metering systems provides new insights and options for power management at LV grids. Due to the mass introduction of DERs, the residual load and maximum power consumption at prosumers households show significant infringements of the contracted value, which might lead to substantial investment costs of grid reinforcement measures for the DSO. In order to limit or avoid such investment costs several options and strategies are currently under discussion. Besides common Smart Grid strategies (where often price signals, centralized controllers and ICT are used) also regulatory changes like system charges tariffs or increased introduction of home energy management systems (HEMS) could provide solutions for this problem.

The work described in this paper is focusing on a scenario where strict power limits are applied to prosumer households. Those prosumer households are equipped with PV, EVs, BESS and HEMS. Also, part of this scenario are end-customers without such systems and options for self-management. Based on the nominal transformer characteristics, a maximum power limit is introduced to all prosumers, which must be fulfilled during each 15-minute measurement interval of the metering system. Each prosumer household is managed individually by its HEMS, focusing on staying within power limits, maximizing the usage of its own PV generation and minimizing possible negative effects on the household owner (e.g. limiting recharging of EV).

The expected results will show the effectiveness of the chosen approach for supporting grid congestion issues at LV levels, as well as its limitations, requirements and possible negative effects to prosumer households.

Integration of Short-term PV Forecasts in Control Strategies of PV Diesel Systems
Submission-ID 315
Pierre Besson 1, Thai-Phuong Do 2, Gabin Koucoi 2, Franck Bourry 2
1 SteadySun, France
2 National Institute of Solar Energy, France

In this work, we will present through simulation, how the integration of short-term PV-forecast into control strategies is an efficient way of lowering operational cost of hybrid PV-Diesel system and increasing PV-rate integration in the system. We will evaluate, using modeling tool, the benefits of different control strategies of the system for different PV-rate integration.


Short-term PV forecast is made using sky images processing techniques. The solution ‘steadyEye’ studied in this work, provides very-short term forecast of the PV plant’s production for the next 60 minutes through cloud mass movement analyzes. For evaluating the benefits of integrating PV prediction in control strategies, a simulation platform called SPIDER, realized by CEA INES, is used. A microgrid system with three gensets, a PV system, and a load is simulated. Energy Management System (EMS) with different control methods, has been implemented. Two approaches are proposed. First, control strategy, based on a 'reactive' operation of the system, is tested: the genset reacts in real-time to power variation of the PV production and the load. A second strategy, is developed and integrates an advanced control with an additional level providing short-term planning using PV forecast data.

Simulations have been performed for a set of 30 days which covers different typical PV daily profiles. Depending on the control strategy used, different figure of merit are investigated such as system stability, fuel consumption, or energy availability (black-out time). These studies are done for different PV-Diesel system sizes and show the benefits of having a control stratgey based on short-term forecasts.


In this work, we have shown how short-term PV forecasts help better operating hybrid PV-Diesel system. Different outcomes can be considered:

- Benefits of advanced control in terms of system operation cost and stability

- PV integration rate can be increased thanks to PV forecasts;

Secure Energy Information Network in Germany - Demonstration of Solar-, Storage- and E-Mobility Applications
Submission-ID 317
Gerd Heilscher 1, Shuo Chen 1, Heiko Lorenz 1, Falko Ebe 1, Christoph Kondzialka 1, Tobias Kaufmann 2, Sebastian Hess 2, Jens Wening 3
1 Ulm University of Applied Sciences, Germany
2 Zenner-Hessware GmbH, Germany
3 meteocontrol GmbH, Germany
INTRODUCTION – The integration of decentralized renewable energy systems into our distribution networks leads to a need of more detailed information about local network structure and state estimation down to the low voltage level. This enforces the transformation of today’s distribution networks into smart grids. Smart Meters, Smart Meter Gateways and Controlable Local Systems are the essential new bricks of the future smart grid. In Germany the new law “Digitalisation of the Energiewende” sets up the rules for network operators to establish this secure energy information system based on the smart meter infrastructure.

INNOVATION – During the last two years the authors developed and demonstrated on laboratory and field level such a secure energy information system. The main innovation of the project is the direct and secure communication with decentralized energy systems such as photovoltaic inverters, battery systems, E-mobiltiy charging stations or power to heat applications within this new smart meter infrastructure, which has been defined by technical rules from the German regulator for data security (BSI). The two way communication is able to read measurement values from the field as well as change setpoints or activate curtailment of decentralized energy systems.

RESULTS - As standardized data protocol IEC61850 is established for every application. The CLS-Gateway translates the local protocols of the different devices (SunSpec, proprietary REST interface, Modbus register,…) into the IEC61850 standard. The CLS-Gateway is connected to the Smart Meter Gateway with a TLS encrypted connection. A CLS-Center hosted by ZENNER-Hessware establishes the communication channels to the CLS-gateways in the field. The CLS-Center also secures the stable, and continues operation of all single communication channels with a ping-response-service. As Backend system the experimental distribution control center (PowerSpectrum5) of Ulm University of Applied Sciences collects the field measurements, activates, and controls the connected energy systems.

OUTLOOK – Based on the actual results of the laboratory and field experiments 500 CLS-gateways will be installed within the SINTEG project C/sells during the next year. With two smart grid test sites at Ulm, further applications like state estimation and power2hear will be tested and the quality of grid operation services for low voltage networks down to each single grid coupling point will be investigated.

Practical Implementation of the SNOOPI-Box for a Smart Voltage Control in the Distribution Grid
Submission-ID 325
S. Hempel 1, J.-D. Schmidt, J.-D. Schmidt, 1, E. Tröster 1, U. Ohl 2, M. Koch 2
1 Energynautics GmbH, Germany
2 EWR Netze, Germany

Decentralized Secondary Frequency Control in an Optimized Diesel PV Hybrid System
Submission-ID 326
Alice Vieira Turnell 1, Peter-Philipp Schierhorn 1, Daniel Masendorf 1, Kateryna Morozovska 2
1 Energynautics GmbH, Germany
2 KTH Royal Institute of Technology School of Electrical Engineering and Computer Science, Sweden
This research aims at investigating if a diesel-based isolated electrical system can be optimized by integrating a high share of solar photovoltaic (PV) generation and if the frequency stability of such system can be improved by including the PV participation in frequency regulation. A case study is developed in order to explore an island's expansion of the installed generating capacity and its optimization. This study uses the tool HOMER Energy Pro to solve the optimization problem and PowerFactory to verify the frequency stability of the proposed system. The PV integration allows for a reduction of diesel fuel consumption, emissions and generation costs. Additionally, in high PV penetration scenarios, the reduced inertia in such systems can lead to high frequency deviations that may trip the system protection. The study demonstrates that the instantaneous frequency deviation after a load and generation imbalance can be reduced by designing the PVs to operate with an allocated reserve and a decentralized time-based secondary frequency control. The system remained stable under different disturbance scenarios with high PV penetration and reduced available inertia, indicating that high PV integration is economically and technically feasible in a small island system.