1
Transmission Investments under Uncertainty: the Impact of Flexibility on Decision-Making Gerardo Blanco, Member, IEEE, Fernando Olsina and Francisco Garcés
Abstract-- Nowadays, higher electricity consumption and need for economic efficiency have led to increased use of the electric power transmission network. After the severe absence of investments in the transmission grid observed in the last decades, the transmission investment problem is currently a topic of increasing interest among the power market agents as well as regulatory authorities. Therefore, tailored investment valuation models are needed for quantifying the contribution of strategic flexibility in the investment portfolios. In addition, models capable of replicating the uncertain evolution of the long-term behavior of power markets represent a reliable benchmark for designing contingent actions against unfavorable unfolding of uncertainty, aiming at ensuring the transmission network adaptation. This paper analyzes the impact of flexibility on the evaluation of transmission investment under uncertainty based on system-wide social welfare. Stochastic simulations are performed in order to characterize the uncertainty behavior of the investment portfolio performance. From these simulations, an appraisal methodology based on a Real Options approach is applied for valuing the strategic flexibility embedded into the transmission projects and finding the optimal timing for investing. The results show how omission or incorrect handling, of ongoing project uncertainty of the key variables could lead to non-optimal decisions. Index Terms- Monte Carlo simulation, Real Options, strategic flexibility, transmission investment, uncertainty
A
I. INTRODUCTION
CCORDING to the Edison Electric Institute, during 30 years, in the Unites States (US), investment in power transmission declined, between 1975 and 1999 at an average rate of USD 83 million per annum [1]. In words of [2], “some of the given reasons are an overbuilt system prior to 1975, retail rate freezes, lack of available capital due to other investment activities, cost-effective generation located closer to load, and uncertain regulation of transmission”. In addition, events which have taken place at the beginning of this century reveal an important requirement of investments in power network. Some of these event are related to: security issues related with terrorism attacks, blackouts of 1996 in US as well as 2003 and 2006 in Europe, , natural gas price spikes G. Blanco is with the Facultad Politécnica at Universidad Nacional de Asunción (UNA), Universitary Campus, San Lorenzo 2111, Paraguay (E-mail:
[email protected]). F. Olsina and F. Garcés are with CONICET at the Institute of Electrical Energy (IEE), National University of San Juan (UNSJ), Av. Lib. Gral. San Martín 1109 (O), J5400ARL, Argentina. (e-mail:
[email protected],
[email protected]).
978-1-4673-2729-9/12/$31.00 ©2012 IEEE
making nukes and coal more competitive and last but not least, the green energy promotion since renewable energy sources are usually remote from the load [2]. Therefore, a limited recovery of the network investments has appeared since the end of the 20th [3]. From 1999 through 2003, transmission investment activity augmented at an average annual rate of USD 286 million, a substantial reversal of trends. Nevertheless, it is not clear what accounted for this trend reversal or whether the change is temporary or longterm [3]. In any case, the theory and tools for valuing transmission investments are still below the requirements of the actual power markets. This issue is indeed true in aspects such as the quantification in economic terms of strategic flexibility [4]. The transmission investment problem is characterized by the nature of the investments involved. Features as: scale economy, irreversibility, low adaptability and intensive capital are present in the investments in network expansions. In addition, the value of an investment is usually amortized over a long period of time and depends on many other developments. Consequently, the transmission investments are strongly exposed to the large long-term uncertainties. Obviously, the unavoidable uncertainties related with transmission investments are better managed with flexibility rather than fixed scenario expectations by the discount cash flow (DCF) approach. Flexibility offers the possibility to adapt the investment plan of the transmission system, rapidly and at a low cost, to any change, foreseen or not, in the conditions that were expected at the time it was decided. [4][5] Hence, strategic flexibility could be seen as a risk management technique [6], which allows properly managing major uncertainties, which are unresolved at the time of making investment decisions. Nevertheless, expressing the value of flexibility in monetary terms is not a trivial task and requires rather refined valuing tools [7]. The Real Option Valuation (ROV) approach provides a well-founded framework –based on the theory of financial options- to assess the intrinsic flexibility of strategic investments under uncertainty [8]. Power transmission investments comprise intrinsic flexibility by multiple strategic options, such as: the option to expand in a posterior stage, postpone the option and abandon the investment in the future [9]. This paper is an extension of the research presented in [10] and addresses the problem of valuation of flexible transmission investment projects on the basis of the social welfare of the electricity market. It proposes a methodology
2
based on Real Options approach for valuing the flexibility of strategic investments in the transport network and finding the optimal timing of the execution of the investment alternatives. In addition a sensitivity analysis of the main uncertainties, which the transmission investments face during the decision making process, is presented. II. POWER TRANSMISSION INVESTMENTS The transmission of electric energy has intrinsic technical features which impact on the investment decisions within competitive markets. These are: • Power flow distributions according to minimal impedance paths, i.e. the flows cannot be guided in accordance with commercial paths; • The thermal limit of conductors as well as the steady-state and transient stability constrains the power transmission; • The flow through each transmission element is affected by any variation of the flow of another grid component. Consequently, a new transmission line could alter power flows in distant and apparently unrelated transmission lines. In addition, transmission investments typically exhibit inherent characteristics which impact on their performance and must be taken into account during the assessment and design of alternative investments [7]: • Transmission investments suffer exhibit substantial scale economy and lumpiness. As is pointed out in [11], the scale economy could lead to underinvestment scenarios and delay of transmission projects. Therefore, expansion investments generally remain operating under their maximum rate during a significant part of their operative life. • Some power market agents, which are not the expansion investors, benefit from the transmission upgrade. Those agents are called free-riders. The public good nature of the transmission infrastructure allows these externalities to appear [12], and induces the agents to wait for someone else to execute the investment. • A substantial fraction of the total required capital must be expended prior to the commissioning of the new transmission line, while the depreciation may take many years or even decades. • Transmission projects are vulnerable to unforeseen scenarios that can occur during the investment horizon. Future demand, fuel costs and technology and location of new generation are highly uncertain variables when evaluating investments. • Transmission network investments, once executed, are considered sunk costs, because these have a nearly zero residual value if the conditions unfold adversely. • In general, opportunities to invest in the transmission system are not of the now-or-never kind. Thus, it is valuable to maintain the investment option open, i.e. wait for valuable, arriving information until uncertainties are partially resolved [7]. Thus, investments in transmission systems have specific features, which must be taken into account when evaluating them. These attributes are crucial in the performance of a particular investment alternative and largely define their
technical and financial feasibility. III. UNCERTAINTIES AND RISK IN THE TRANSMISSION INVESTMENT ASSESSMENT PROBLEM Nowadays, power systems require strategic decisions to invest in highly uncertain environments, characterized by the unknown of relevant aspects such as: evolution of demand, location of new generation, fuel costs, discount rates, investment costs of transmission projects, availability of system components. Therefore, there is currently a broad gap between what managers would want to assess and the capabilities of tools and available information to conduct it. Consequently, uncertainties play a crucial role within the transmission investments assessment, and accordingly, their behavior should be properly replicated within the appraisal models. The problem of valuing transmission system expansions is better suited to be addressed as a risk management problem [13], by formulating a transmission expansion plan that led the planner to make changes in a simple, economic way for adapting the plan to the unfolding of the uncertain variables. Currently, the risk analysis theory is broadly applied by planners who try to cope with investment decisions under uncertainty, due to the fact that it constitutes a consistent and systematic method for structuring the decision making process. The traditional approach to project evaluation uses the method of Net Present Value (NPV). This appraisal approach presents good performance when the project is exposed to low or null uncertainty in their state variables and does not have any strategic flexibility option (option to defer, expand, switch, abandon, etc.). Hence, economic evaluations of the plans for expansion of transmission systems, based on the traditional method of NPV, do not value the ability to conduct a proactive risk management that offer each investment alternative. That is why an investment, which characteristically has the option either to expand at a later stage, to defer certain decisions or to abandon investment in the future, etc. should consider the monetary value added by this flexibility in its assessment. This value cannot be incorporated by the traditional NPV analysis. In these cases, the most appropriate tool for evaluating investments would be the evaluation of Real Options [14]. The methodology of Real Options is not limited only to determine the optimal time to invest, but also allows quantifying the flexibility to adapt the executed investment to unpredictable events, such as the evolution of uncertain variables to scenarios considered unlikely. Once the new information arrives, the investor may decide to expand, reduce or simply cease the operation. These flexibility options are quantified according to the volatility of the evaluated investment returns. Thus, if uncertainty of these returns increases, the value of flexibility options will increase, and therefore, the need to take them into consideration turns indispensable.
3
Hence, by means of the optimal execution of the flexibility of investments, the value of investment projects can increase with increasing uncertainties. Then, the key is flexibility in dealing with the uncertainties by having various options in place that can be exercised as new information emerges. The options have the source of their value in the fact that they set a floor under possible losses of the investment. IV. STOCHASTIC DYNAMIC PROGRAMMING BASED ON EXPECTED PRESENT VALUE (DPE) The value of the flexible investment portfolio is estimated by finding the optimal exercise time of the flexibility options. Dynamic Programming is a suitable tool for conducting this task. This approach could be structured graphically as a decision tree, and the real option compels a search of the optimal timing to invest and hence determines the optimal decisions to make. At a generic time t, the model will estimate the expected NPV of the investment portfolio considering a wide likelihood of the uncertain variables under two scenarios: either invest now or hold on the investment until the next period [15]. For the sake of clarity, the starting point of the analysis is the assessment of an investment portfolio in the transmission network with the option of deferring the investment. Assuming that the project's license has T years to expiration and the capital required for investment in the year t is I(t), the expected value of the investment project is considered as the underlying assets. The risk-free rate is denoted by r. The optimal policy of exercising the options is derived by comparing the intrinsic value of the deferral option with the value of keeping alive the option using backward dynamic programming techniques. The problem starts from the latest year and working backward is completed in the first year. For the last year, the decision problem is modeled as: Exercise, if V (T ) ¢¡ PV¯±° I (T ); (1) Do not exercise, if V (T ) ¢¡PV¯±° b I (T ) Correspondingly, the optimal rule at the year T is: (2) V * (T ) max ¡ ¡¢ PV ¯°± I (T ) ; 0¯° ¢ ± In any year 0
