Recent Advances in Intelligent Control, Modelling and Simulation

An Investment model for the Smart Grid BAYASGALAN TSETGEE, FABRIZIO GRANELLI Department of Information Engineering and Computer Science (DISI) University of Trento Via Sommarive 5, 38123 Povo, TN ITALY [email protected], [email protected] Abstract: - The paper discusses next-generation energy efficiency optimization enabled by incorporating smart grid technology into the current electric grid system and exchanging real-time power information between the supplier and the consumer. In this paper an investment model is introduced for development of the smart power grid management system, that focuses on developing new services by activating demand response market system, power transactions, investment allocation etc. through a real-time pricing. Key-Words: - Power pricing, reactive power, investment model, prosumer The paper is structured as follows. In Section 2, a brief description of the reactive power management is presented. In Section 3, the investment model for establishing smart grids is described in detail. Section 4 describes the risk problem. Finally, conclusion summarizes the contribution of this work with respect to the objectives of the future work and is drawn in section 5.

1 Introduction

Electricity markets have become very complex, leading to a wide range of new opportunities for power producers to optimize their productivity. The main tasks power producers have to face include production planning and optimal pricing. Sources of stochastical fluctuations, such as high volatility of energy prices and the uncertainty of the production, are becoming increasingly relevant. In this scenario, it is a requirement to design the appropriate investment and pricing model that could better capture the uncertainties inherent in the electricity grid and would encourage investments in new smart technologies. Smart grid investment model calculates the impact of different smart grid investments, along with their strategies and evaluate the costs, and track its benefits across the distribution spectrum. It provides: • smart grid investment analysis in terms of costs and benefits parameters and forecast their impacts; • guidelines on smart grid implementation program, and cost-effective strategies; • opportunities for new products, services and markets; • power quality and generation storage options; • asset optimization with real time monitoring and achievement of operational efficiency. The novel model presented in the paper is based on a decomposition of the investment model of energy and power in smart grid.

ISBN: 978-960-474-365-0

2 Power management

Reactive power price issue has gained considerable attention in the price based competitive electricity markets. Along with the power pricing, with the growing interest in determining the costs of ancillary services needed to maintain the quality of supply, the reactive power price has also gained great importance. Many approaches and models for determining spot pricing, marginal price, load flow, production cost, problem of spinning reserve pricing, congestion alleviation cost, security components, dispatch controller, FACTS controllers on transmission pricing and cost model have been proposed in [1-10]. In [5], the impact on nodal prices has been determined considering three different reactive cost model for generators’ reactive power cost models. The impact of FACTS devices have been incorporated with their cost model. There are tree methods for the evaluation of the reactive power cost: • Triangular approach • Maximum real power based approach • Maximum apparent power based approach.

57

Recent Advances in Intelligent Control, Modelling and Simulation

generator are taken into consideration for the minimization of the marginal cost. While the TSO is responsible for maintaining the power balance on a momentary base, the participants have the obligation or the target to balance their position over a period of time, also known as the program time unit. Generation costs are minimized while satisfying the energy equality constraint by producing at a constant level during the full program time unit. The power system’s reactive power management system calculates the reactive power by utilizing the voltage management system central control unit and reactive power distributor and on-site information acquisition device to measure the system data in real-time and performs the functionality of reactive power management by adjusting the power plant AVR and StatCom by using the above results [12]. It improves the quality the in electricity distribution services by minimizing excepted generation costs and imbalance charges, and helps TSO to manage security of electrical network.

The ancillary services market works in a “price as bid” mechanism. This means that bids from market participants are dispatched at the proposed price according to an economic order (lower selling bids/higher purchasing bids are activated first), until the total demand is covered [13-15]. The cleared power price is calculated by the sum of energy quantities allocated to bilateral contracts, day-ahead and ancillary services markets, respectively, with the corresponding prices. And it depends from power production and communication cost. ) Cleared price model requires an evolution of bid prices to generate the relevant price scenario, which will be exploited by the optimal control policy. The price parameters can be determined by the econometric model using the dynamics of electricity prices. Energy can be assimilated to a particular traded stock, whose price is affected by high volatility and unpredictability. The balancing market – a selection of suppliers of electricity in real time, based on minimizing the cost to meet the emerging demand for the actual system conditions, based on the bids of suppliers of balancing electricity (generators and consumers with variable load) on the partial content and/or unloading compared with the plan, and involves the formation of nodal equilibrium price for balancing electricity for both buyers balancing electricity and for its suppliers [11]. Market participants, consisting of generators and loads to trade electricity on the power market in order to satisfy loads, earn profit and contribute to the power balance in smart grid power system in energy management system based on the exchange real-time information between power suppliers and consumers to optimize the supply and demand of power. Some energy producers, such as prosumers are uncontrollable and have different behaviour that cannot be expected or predicted beforehand. So, it is necessary to develop extensions of the optimal bidding problem to consider a combined production unit, consisting many generators, with different characteristics and feeding (coal, hydro, wind, solar), thus leading to dynamical modeling of the plant behaviour, subject to ramp-up/down constraints. Total generation equals total load at any time. The actual balance management is done via the imbalance settlement system of the Transmission System Operator (TSO). Minimum and maximum power setpoints and ramp rate constraints of each

ISBN: 978-960-474-365-0

3 Formulation of Dynamic investment model

Development of the electricity market requires innovation and investments on devices for smart grid technology, that increase its efficiency in its use of power generation facilities and in its delivery of uninterruptible and high quality consumer-centered power services. Establishment of the smart grid will not only solve future energy problems and improve the quality of life, but will also bring other benefits such creating new jobs, reducing greenhouse gas emissions, lowering dependence on energy imports, increasing exports, creating domestic consumption, and avoiding the need to build new power plants. Smart Grid Establishment and Usage was enacted to promote the development and usage of a stable and systematic smart grid, to actively cope with global climate change by establishing the legislative or regulatory grounds for fostering the related industries, and to contribute to innovation of the energy usage environment and development of the national economy by establishing a low carbon green growth-oriented future industrial base. At limited opportunities for investment development the following should be taken into account: • establishment of regulatory basis for the smart grid, and utilization and protection of energy

58

Recent Advances in Intelligent Control, Modelling and Simulation

information and securing of reliability, stability, and interoperability of information; • development of the grounds of financial support for the establishment of smart grid infrastructure, and its systematic designation; • promotion of smart grid international cooperation, PR, business support and export; • management comprehensively power IT and smart grid R&D project; • technology development; • power industry structure; • demand management, billing system; • metropolitan investment financing mechanisms, including – reinvestment (domestic and foreign investment capital, implementation, timing, and financial sharing by feasibility review etc.) In market conditions of optimality criteria of energy development are purely economic criteria: the maximization of profit, and return on investment, etc. Given the limited availability of investment funds from external sources, as well as the possibility of accumulation of equity in the form of reinvested profits, to define and justify the sequence of construction and putting into operation of objects. Dynamic investment performance criteria should include a reasonable procedure discounting. Socio-economic assessment options for the smart grid system development are given over a period of time with the reflection of the formation of these estimates over time. An important role is played by the procedure of comparing and aligning costs and benefits occurring at different times to a single point in time. In this dynamic optimization it is necessary to consider energy development and to evaluate current costs and investment in new construction and renovation of the results of its operation for the period of power development and output it to the optimum mode of operation. Let's consider the following dynamic model. sectors of electricity market can be divided into segments or objects, which are subdivided into new construction objects or modernization (reconstruction) objects. Let's enter the following notations: – total amount of investment attracted on development electricity market; – share of aimed at the development of ); sector ( - volume of investment in sector ; – investment into new/reconstructed objects in sector in the period ; – power setpoint of new/re construction object in sectors in the period ;

ISBN: 978-960-474-365-0

– cost of new/re construction object in sector in the period ; – power setpoint in sectors , realization existing at the beginning of the period ; – maximum power setpoint of new/re construction object in sectors ; – total investments from allocation source l in the period ; – price of investment from allocation source l (credit percent); – maximum receipt of funds on investments from l allocation source in the period of ; – balance of investments of new/re construction objects in sector at the end of period (therefore, ); – demand of electricity generation/power from producer in the period of ; – power price/profit of generator/producer during ; – normative load in sectors of producer that corresponds to the offered standard determined by indexes ( ); – reduction coefficient of expenses/profit sum by realization beginning time: in period where – time horizon, which have passed since the beginning of realization until the end of the period ; – number of the periods into which realization term is divided; – interest rate; – normative coefficient of payback on investment profit in electricity market : – payback period in average in electricity where market on investment profit; – time horizon in period ; – assignments from profit on reinvestment in period ; – tax assignments in shares from balance profit; L – set of investments sources; lc – set of investment source from own accumulation. The dynamic optimization model is formulated as:

Balance of power by sectors:

59

(1)

Recent Advances in Intelligent Control, Modelling and Simulation

funding development in stages and source of funds, as well as the optimal structure and dynamics of directions. Solution series of the problem in the previous stage are calculated power input objects resulting from new construction and reconstruction of facilities:

(2) Balance of investment by power reconstruction:

Restrictions on the amount of construction work of existing facilities:

(3) new/re

(10) In (2) these additional facilities are added to the stage to the initial power (phase ) objects of this sector:

(4) Balances on the formation and use of investment areas:

. (11) In (4) keep in mind that reserve power is reduced by objects already spent and reconstruction in the previous phase:

(5) Investment sources: a) By fund investments generated from their own savings in previous periods:

b)

c)

(12) We can decompose the model as a system of interrelated models in stages to the following terms and obtain a model for an arbitrary phase , given the results of (11)-(12) for the system development for the preceding stages .

(6) According to other investment sources (from the state budget, local budget, foreign investments, etc.): , (7) On receipt of the investment balance (unused funds at the end of the period) after the period τ :

(13) In model where – time horizon in period , as income and costs are reduced to the initial year of each period, and their income is uniformly distributed within the period. The discount factors are meaningful reduction coefficients: they cite the effects obtained for the time horizon (assuming they even produce income for the time horizon to the start time of each period ). As a fixed common factor in the objective function, it does not affect the choice of optimal development and functioning of market on stage . However, the resolution of any problems with the above calculation for periods effect plays an important weighting effect at each stage. Usage of their own savings as a source of investment necessitates as early as possible to make full use of the funding for development of profit and

(8) Variables: (9) The proposed model of investment attractiveness, based on the balance of demand, existing and new capacity, can be used to determine the volume of necessary investments and their temporal structure by sector and facilities, the

ISBN: 978-960-474-365-0

60

Recent Advances in Intelligent Control, Modelling and Simulation

return on investment at the previous stages. This forms the basis for own savings while separately maximizing the return on investment at each stage. Restrictions on fund investments generated from their own savings consider the reinvestment funds from profits generated in the previous period. Therefore, by solving the phase for the formation of the task steps value is calculated:

the reduced value of the net increase in assets for the period :

(17) Optimal values indicate the optimal structure and dynamics of power different directions. It's on standard, which can be implemented in stages of the period, based on the types and proportions. Optimal values determine the allocation of investment sources by stages. Values define the detailed investment program: the direction of funds for new and reconstruction objects in each sector, as well - the temporal structure of the investment program (index ). Optimal values show the necessary sequence (index ) of new and reconstruction objects in each sector. Substituting the optimal values of the above variables in the constraint the milestone tasks, we obtain the corresponding balances on capacities and resources justifying the optimum realization of its development ( ). Addressing global challenges requires a largescale influx of mid-term investment and the creation of an effective mechanism to attract them and use both at the state and among the subjects. The need for organizations to invest in energy sector causes great practical need for research investment process (e.g. in Mongolia) and the theoretical justification for choosing the directions of investment activity.

(14) Deductions from the balance profits to pay interest for the loan are determined on the basis of data on investment for the previous years. Factor means that the maximum average profit obtained in step , should be multiplied by the time horizon in the period. It is up to the sum of profits and deductions are taken to increase production in the subsequent period. Investments for new/re construction projects should not exceed the amount of external sources of the period and accumulated by the time funds. Proposed decomposition model into subtasks to be solved sequentially on temporary stages, fundamentally solve the problem of dimensionality. Dimension of the programming problems at each stage are reduced in time . As a result, the solution series of tasks following results for the period

The carrying amount of the reduced earnings for the entire period is determined by the formula:

(15) The numbers in square brackets are the sum of the balance sheet profit derived from the values of the objective function in an optimal solution for each stage , excluding interest payments on loans and reduced amounts of investment. The reduced amount of net income for the period can be calculated taking into account the interest payments on loans and tax deductions:

4 Risk problem of investment

In terms of transformation priorities in assessing the status and prospects of the investment scope, particular importance is attached to the structure of performance criteria in investment demand, efficiency and payback period of investment. For the analysis of investment projects, the following selection criteria: • investment indicators: the net present value (NPV); • internal rate of investment (IRR) and return on investment; • financial performance: the coverage ratio, the amount of fixed assets and value for money; • risks: financial, political and economic;

(16) If the period has duration of not less than the standard term return on investment, the costeffectiveness of development can be identified by

ISBN: 978-960-474-365-0

61

Recent Advances in Intelligent Control, Modelling and Simulation

• time criteria: time periods of the project, and return on equity turnover. The complexity of investment decisions due not only to the variety of sources of funding, but also features investments, which include: • risk degree of the planned income; • relatively long payback period; • existence of a critical area of economic return of investment, for ex. zone does not guarantee the stability of the profit . We can solve the problem of determining the share of investments to reduce risk on the basis of data . Obviously

(21) Income is stochastic, it can be assumed that any particular value return is the realization of a , dispersion random variable with mean . The covariance of (variation) two random variables is calculated as . If the variation of income is zero, then there is no uncertainty, and consequently no risk. With the growth of variation and uncertainty is increasing profitability, hence the risk of receiving less income. The value of a random variable may deviate unfavourably by the expected value ; as a measure of the amount of risk we take a variation: . The expected revenue is

(18) Statement of the problem of minimizing the risk function:

In following conditions:

(19) deviation from the expected value:

(20) dispersion of income:

Where - investment risk in object in sector . In the financial analysis to select the most efficient portfolio use preference curves, reflecting the investor's attitude to risk. The investor determines its preference curves presented in the form of two-dimensional plots of the return from the risk of loss. The problem of optimizing the investment portfolio is to determine how much of the portfolio should be set aside for each of the investment in such a way that the value of the expected return and risk optimally meet the objectives of investors. Investor's goal is to minimize the risk of the portfolio, where risk is measured by the variance of the portfolio. The investor usually sets limits as to the manner in which a portfolio can be constructed. For example, the objective function may be to minimize the risk for a certain level of income, as well as restrictions on the minimum and maximum share that can be invested. If we denote the expected sector , the total income of the investor or the utility function is defined as:

ISBN: 978-960-474-365-0

(22)

(23)

(24) where – covariance matrix of specific abnormalities. As a result, the problem reduces to the problem of minimizing a quadratic functional - variation of income:

In conditions:

(25)

(26)

62

Recent Advances in Intelligent Control, Modelling and Simulation

[7] M. Cerullo, G. Fazio, M. Fabbri, F. Muzi,

(27)

G. Sacerdoti - Acoustic signal processing to diagnose transiting electric-trains. IEEE Transactions on Intelligent Transportation Systems, Vol. 6, No. 2 June 2005. [8] F. Muzi, C. Buccione, S. Mautone – A new architecture for systems supplying essential loads in the Italian High-Speed Railway (HSR). WSEAS Transactions on Circuits and Systems, Issue 8, Volume 5, August 2006. [9] F. Muzi, L. Passacantando – Improvements in Power Quality and Efficiency with a new AC/DC High Current Converter. WSEAS Transactions on Circuits and Systems, Issue 5, Volume 7, May 2008.

This problem is a quadratic programming problem, for which several efficient solvers exist.

5 Conclusion

The presented mathematical model ensures the feasibility, the reliability and the efficiency of the future smart power system in emerging markets (e.g. Mongolia), by anticipating and supporting market based operation and decentralized decision making, which allows assessing the prospects for supporting a national smart grid roadmap. The paper aims to contribute to the improvement of the quality of people’s lives and to spur national economic growth, by supporting the smart grid business, which is the core policy of low carbon and green growth, and by nurturing the smart grid as a new growth engine through green energy technology innovation.

[10] C. Buccella, C. A. Canizares, C. Cecati, F. Muzi, P. Siano, Guest Editorial for the Special Section on Methods and Systems for Smart Grids Optimization, IEEE Transactions on Industrial Electronics, Vol. 58, Number 10, ITED6, October 2011. [11] Zagdkhorol, B.; Tsetsgee, B., "Modeling of balancing market for competitive wholesale electricity market," Strategic Technology (IFOST), 2011 6th International Forum on , vol.1, no., pp.609-613, 2011 [12] Jang Moonjong, B. N. Ha, S. W. Lee, D. Y. Seo, “The Study on the Construction of the Smart Grid Test Plant and the Integration of the Heterogeneous Systems”, WSEAS Transaction on Power Systems, Issue 2, Volume 8, April 2013 [13] Zagdkhorol, B.; Andreyevich, S.V.; Vitalevich, S.S., "Development of the competitive electricity market in Mongolia," Strategic Technology (IFOST), 2013 8th International Forum on, vol.2, no., pp.545-552, 2013 [14] Zagdkhorol, Bayasgalan; Tsetsgee, Bayasgalan, "Real-time balancing services in the electricity market," Strategic Technology (IFOST), 2012 7th International Forum on, vol., no., pp.1-4, 2012 [15] Zagdkhorol, B.; Tsetsgee, B., "Score for the model of Mongolian competitive electricity market and the possibility of improving its functioning," Strategic Technology (IFOST), 2010 International Forum on, vol., no., pp.216219, 2010

References: [1] Laura Puglia, Daniele Bernardini, Alberto Bemporad, “A Multi-Stage Stochastic Optimization Approach to Optimal Bidding on Energy Markets”, 50th IEEE Conference on Decision and Control and European Control Conference (CDC-ECC), 2011, pp.1509-1514. [2] Jasper Frunt, Ioannis Lampropoulos, Wil L. Kling, “The impact of electricity market design on periodic network frequency excursions”, 8th International Conference on the European Energy Market (EEM), 2011, pp.550-555 [3] P. Patrinos, D. Bernardini, A. Maffei, A. Jokic, and A. Bemporad, “Two-time-scale MPC for economically optimal real-time operation of balance responsible parties,” 8th IFAC Symposium on Power Plant and Power System Controls, Toulouse, France, 2012. [4] Antonio J. Conejo, Juan M. Morales, Luis Baringo, “Real-Time Demand Response Model”, 2010 IEEE Transaction on Smart grid, vol. 1, no. 3, pp. 236-242. [5] Kumar, A.; Wenzhong Gao, "Social welfare based nodal pricing with FACTS cost models in hybrid electricity markets," Innovative Smart Grid Technologies (ISGT), 2012 IEEE PES , vol., no., pp.1,8, 16-20 Jan. 2012 [6] Muzi, F.; De Lorenzo, M.G.; De Gasperis, G., "Intelligence Improvement of a "Prosumer" Node through the Predictive Concept," Computer Modeling and Simulation (EMS), 2012 Sixth UKSim/AMSS European Symposium on , vol., no., pp.311-316, 14-16 Nov. 2012

ISBN: 978-960-474-365-0

63