[ieee 2012 ieee international conference on smart grid engineering (sge) - oshawa, on, canada...

IEEE International Conference on Smart Grid Engineering (SGE’12), UOIT, Oshawa, ON, 27-29 August, 2012

P<SGE12-OS-5-Paper 15>-1

Japanese Test Facilities for Smart Grid

M.Marmiroli, Member, IEEE, M. Koshio and Y.Tsukamoto

Mitsubishi Electric Corporation Abstract The Japanese government policy for carbon emission reduction is based on the increasing of generation capacity for photovoltaic to 28GW by 2020 and 53GW by 2030. Even if renewable energy sources are expected to contribute to the emission reduction, there may be some technical difficulties to integrate a large amount of renewable sources to the existing electric power system. Difficulties and challenges associated with the renewable energy sources are mainly related to the location of the sources and to the unstable output of the generation. Smart grid technologies are the key to solve these challenges. This paper introduces a smart grid test facility developed starting from 2010 in Amagasaki Japan. The aim of the facility is to create an advance environment with a large amount of renewable sources integrated. In a Mitsubishi factory, 4MW of photovoltaic panels are connected to a distribution grid in several combinations and voltage level. The test facility is used to develop and test new algorithms, systems and equipments for the smart grid of 2020. The paper focuses on equipment and technologies to ensure high power quality in the power system especially regarding frequency and voltage stabilization. Keywords: photovoltaic integration, voltage control, frequency control, testing.

1 Introduction A massive increase of installations of photovoltaic, wind generation and other renewable

energy sources is promoted by governments and agencies as an alternative to fossil fuels around the world. Even if renewable energy sources are expected to contribute to the emission reduction, there may be some technical difficulties to integrate a large amount of renewable sources to the existing electric power system. And further, future renewable integration has enormous potential to cause a paradigm shift in electricity supply business.

In addition to the environment issues, recent events such as the East Japan Great Earthquake in March 2011 and the consequent nuclear accident of Fukushima nuclear power plant highlighted the importance of a robust and redundant power system.

Smart grid solutions are the technologies that combining the electrical engineering and the information and communication engineering allowed increasing the readability of the power network and at the same time to improve the condition for renewable sources penetration.

Mitsubishi Electric Corp. (MELCO) is constructing a test facility to promote the R&D of smart grid technologies, which could contribute to the stable and efficient supply of electricity and also contribute to develop and evaluate the new business model in power sector. The facility is composed of new generation equipments such as mega-solar, batteries, and new power devices such as small size of SVC, new types of power conversion systems. Many technical developments for powers system dispatch and control such as advanced EMS/SCADA, DAS and AMI, will be tested and the effectiveness of algorithms and functionalities will be proved. This paper presents the in-house facility at MELCO, which is now under construction to promote the research and development of smart grid technologies. The future electricity supply system could be simulated and analyzed on the facility.



2 Technologies for Smart Grid To overcome to the previous described challenges, several technologies may be developed and

applied to the power grid operation, in order to obtain a smarter grid, able to dynamically adapt to the changes in demand and production sources. -Forecasting of photovoltaic and wind power production A precise estimation of the long and short term production of renewable sources and of the

electric demand is fundamental to obtain a cost efficient and risk free operation plan for the conventional production resources. Based on the weather forecasting and historical production data, it is possible to develop a detail forecasting for renewable sources production composed not only of the most probable value but also a band of possible values with high expectation. The data can be feed in a risk/cost minimization unit commitment algorithm to determine the operational plan of conventional units that satisfies the required reserve and regulation constraints. -Voltage control for distribution network In a conventional distribution network the voltage stability is ensured by the operation of the

voltage tap of the transformers and in the network design. To avoid voltage fluctuation in presence of high penetration of distributed generators, an expensive and not realistic solution maybe the redesign of the network. Fortunately, power electronics equipment such as SVC (static Var compensator) or SVR (step voltage regulator) and PCS (power conditioner system) of the photovoltaic systems can be operated in order to maintain the voltage in the network optimal. These equipments may be controlled locally, if the operation value is decided by the voltage value at the equipment itself, or centrally. An optimization algorithm that minimizes losses in the network can decide the control value for SVR, SVC and Var compensation from the PCS of photovoltaic. -Storage devices control Storage devices, such as batteries, connected to the power grid improve the reliability and the

efficiency of the power system operation. Batteries can be utilized to mitigate the voltages problems if connected to the distribution network and controlled for both active and reactive power. Due the high speed in output change, batteries may be utilized for frequency control and load

following. In case of large energy capacity, they can also mitigate the overproduction problem. If storage devices are distributed in the network, coordination among them is fundamental to

avoid increasing of charging and discharging losses. In the future, it may be forecast that, battery devices in electric vehicles will be partially

utilized to support grid operation. -Information and communication technologies Forecasting, central voltage control and optimize storage device control are possible only if a

large amount of real time data and historical date is available. For this reason, ICT (Information and communication technologies) is the key technology for smart grid. The communication network has to be differentiated and flexible. Expensive and high reliable optical fibers and microwave may be implemented to acquire real time data for storage device control; economically competitive radio frequency or PLC can be utilized to monitor distribution network devices or for smart meters. Smart meter is a key device in the next generation grid. Through the communication network,

the smart meter is not only the device that allowed the consumer to understand the usage of electricity and optimized it implementing tariffs and incentives, but it can also be the source of quasi-real-time data that may be useful for forecasting and voltage control.



3 Test Facility Overview In 2010, MELCO decided that all the technical solutions and devices for the smart grid needed

to be tested all integrated together in order to understand the interoperability of the grid. Fig.1 shows an overview of the facilities.

Figure 1. Test Facility Overview

Equipments in the facility include: - 4MW of photovoltaic on the roof of almost all buildings in the site connected with PCS of 4,

10, 50, 100, 250 and 500kW of capacity with Fault Ride Through (FRT) functionality and remote active and reactive power control - Large storage devices: 500kW of NaS battery 250kW of Lithium-ion battery and 200 kW of

Ni-H battery. All the devices can be remotely and locally controlled both in active and reactive power output - 200 kW and 250kW motor-generator devices connected to a fly-wheel to simulate a hydro

pump-storage unit and synchronous generator. - A 8 km distribution network (6.6kV) that can be freely extended to 15 km with the control of

distributed impedance among the network. - 420kW of programmable load that can be connected or disconnected following the needs of

the tests that have to be performed - A digital simulator connected to the site power system through a BTB (back to back) creates

critical situation such as frequency fluctuation and voltage drops or sags, without any effect to the commercial grid. - Distribution network devices such as SVC, SVR, network switches equipped with sensors

and communication interface. - 150 of smart meters to monitor on-line demand, voltage and current at the low level voltage

of offices and programmable load. - An EV high speed charge station and a low voltage charge plug where three electric vehicles

can be charged. - Three communication networks: optical fibers for balance control (mainly battery and

generator devices), metal based OFDM (Orthogonal Frequency Division Multiplexing) for voltage control for distribution network, and radio frequency mesh network for metering. Mitsubishi Electric developed also a model house to demonstrate the importance of the



demand side not only at an industrial level but also at a residential level. The smart house is a low consumption building equipped with solar system, charge/discharge capability electric vehicle, home energy management system and smart electric appliances. The test facility was develop with the intent to have a real analog power system to simulate

and verify the performances of algorithms and equipment under the following conditions - Sever power system conditions (earth fault, short circuit, generator fault - Political changes (liberalization, interconnection requirements, wheeling rules) - Changes in business environment (power system management, regional distributed resources). - Climate change (temperature, humidity, solar radiation, wind, etc.) Mitsubishi Electric is now testing and demonstrating the feasibility of the following: (1) power supply and demand balance with high penetration of renewable energy (2) distribution voltage stability in case of a large amount of distributed generators (3) power-saving and energy conservation (4) blackout prevention and outage time reduction (5) demand response value in severe power system condition (6) testing of equipment before commercialization

4 Controlling the Smart Grid To demonstrate the operability of a power grid with distributed generation resources, the test

facilities are operated with a multi-layer monitoring and control energy management system. Since it is possible to simulate a full utility grid with the digital simulator, a wide area energy management system operate the virtual power system calculating the total area renewable energy production forecast, the demand forecast and based on these data the day ahead unit commitment, the economic load dispatch and the frequency control. The wide area EMS is interfaced to the regional energy management system that monitors and controls an area inside the utility. In the test facility in addition to several virtual areas there is a physical area that is the factory itself with 35MW of load and the other equipments previously described. The regional energy management system is modeled as a big battery with constraints in the wide area EMS. The EMS sends signal to increase or decrease the total production/consumption in the area based on the situation of the total grid and the regional energy management system, dispatch batteries, controllable loads and distributed generators to perform the requests of the EMS. A distribution management system monitors and controls the voltage in the distribution network operating SVR, SVC and PCS. The Metering Management data are used in the distribution management system as input for the state estimator module. The demand side management systems (BEMS and HEMS) communicate both with regional management system and metering system.

Figure 2. Proposed Operation System



5 Smart Grid related Technologies A. MicroGrid The expectation for regional power grid (Microgrid) with renewable energy resource is to

adapt total energy output in response to changes in weather and demand. Such a system would reduce the impact that PVs and WTs have on commercial grids and allow the interconnection of more dispersed energy resources (DER). This is a new concept of introducing DERs, and in particular renewable energy resources (RES). In the microgrid, all the power devices and load are linked to the independent power grid, and

several thermal generations and consumption equipments are connected in the heat network, and further all the devices are communicated in the high speed communication network. In case of such a small independent grids with the unstable output and fluctuated load inside, it is difficult to keep the certain level of power quality. B. Smart Metering The infrastructure for smart metering is essential for the next generation for electricity supply

industry. Many kinds of services on the infrastructure are expected. Not only interval and on demand meter reading, and remote connect/disconnect, but also outage management, flexible tariff, and demand response are typical examples. The infrastructures are composed of three parts; meter access system, backbone network system, and head end system. The meter access system includes meter device itself, communication unit and utility gateway. For the future service requirements such as demand response, communication unit have to communicate not only utility gateway but also home gateway inside individual home. The fiber optics and mobile network are the typical composition for the backbone network. The composition in the backbone network is different of utility companies and locations, however high reliability for the system would be highly required for the future. The head end system is composed of data collection, meter data management and network

management functions. That is a large scale network system, in which the number of ten millions meter should be covered and managed. This large scale head end is the mission critical and many service functions make some access. The formalization of access network is very controversial issue from the point of view of

which kinds of media should be adopted and how the network should be broadened. The wireless mesh is one of the most effective technologies in Japanese metering situation. MELCO developed the light weight mesh network middleware for smart metering infrastructure. This middleware embodies high level of scalability, self-healing /redundancy feature, and applicability on several kinds of radio frequency.

Figure 3. Smart metering Infrastructure



C. Distribution Management Generally, electricity is transmitted from electric power plants via substations to users.

However, the large amount of the roof-top solar generation makes the distribution flow be reverse. This reverse flow can make the voltage fluctuate dramatically on a minute-to-minute basis depending on local weather conditions. In Japanese, voltage regulation on the distribution line is 101V±6V at the connection point of customer. MELCO developed a voltage control system for the future emerged environment as previously described, which incorporates optimal power flow computation software. This system is executed as one of sub-system of distribution automation system and has been designed to analyze electricity flow to ensure the appropriate voltage in the high computation time. In order to manage the demand and quality of supply, the distribution automation system

cooperates with AMI (Automatic Metering Infrastructure), FEMS (Factory Energy Management System), BEMS (Building Energy Management System), and HEMS (Home Energy Management System) via a high-speed communication network.

6 Conclusion In this paper, MELCO project for smart grid was introduced. A facility that will allow the

development and testing of new products for smart grid is under development. For increasing the flexibility of the facility a digital simulator is integrated with a distribution grid equipped with several power electronic devices and distributed sources. Several communication networks are also available to allow testing of both electrical and communication interface of new products.

7 References [1] METI, Report of Evaluation Committee for a low carbon power supply system “For the

development of a low carbon power supply system” July 2009 (in Japanese) [2] Tsukamoto Y. “Developing new technologies for the next generation power system” The

Journal of Construction, December 2010 (In Japanese) [3] Marmiroli M. et al “Development of test facilities for next generation grid” Bologna CIGRE

Symposium September 11-13, 2011



Authors

Marta Marmiroli received her degree in Nuclear Engineering in 1996 from the University of Bologna, Italy. After a period of study at Tokyo University she joined Mitsubishi Electric Corporation in the 1997 working mainly on power market design and market supporting software. In 2008 she received the Ph.d degree from Waseda Univeristy in Energy and Environment. Her research interests include power system planning and energy economics. She is a member of IEEE and IEEJ. Masanabu Koshio received his university degree on Physics from Tsukuba University in 1992. After graduation he joined Mitsubishi Electric Corporation where is in charge of the smart grid test facility project and technologies development for smart grid. He is a IEEJ and CIGRE member. Yukitoki Tsukamoto received his M.S. in electric engineering from Waseda University, Tokyo, Japan in 1990. In the same year he joined Mitsubishi Electric Corporation as engineer. He is currently deputy general manager of the Power System Engineering Group. His research interests include energy economy, deregulation of the power system industry, assets management, and smart grid. He is a member of CIGRE and IEE of Japan.

[ieee 2012 ieee international conference on smart grid engineering (sge) - oshawa, on, canada...

Documents