Abstract

Security management of a transmission network has always been one of the challenging tasks for system operators especially within a competitive environment where a transmission network is commonly operated close to its limits. With this regard, Total Transfer Capability (TTC) was introduced by NERC in 1996 to be used as an index which measures the ability of a transmission network to carry or move electric power from one location to another across the system. It also represents one of the system operational limits, useful information for system operators in managing the security of a transmission network. There are several issues to be considered when evaluating the TTC, i.e. the uncertainty associated with the system parameters and conditions, voltage and transient stability constraints.

Over the last few years, it can be witnessed a rapid renewable energy development throughout the world mainly as a result of rising environmental concerns. Renewable energy is expected to be a clean and sustainable energy source for future electricity production. Japanese government already set out a target of 6610 MW wind power and 53 GW photovoltaic (PV) by 2030. However, a large penetration of renewable energy may pose an undesirable impact on the system security and reliability. Unlike conventional energy sources, renewable energy sources have very limited dispatchability and intermittent power output fluctuating with the wind speed or climate. In the past, the impact of the uncertainty associated with the renewable energy was minor and often neglected. Nonetheless, as more and more renewable energy is penetrating into a power system, the impact becomes significant and cannot be simply overlooked.

Another technical issue for the TTC evaluation is the integration of the voltage and transient stability constraints. From the system security viewpoint, the voltage stability should be ensured by operating a system with a sufficient margin away from the voltage collapse point. In many power systems, the transfer is sometimes restricted by the transient stability constraint following large disturbances. Some studies include the transient stability constraint into the TTC evaluation using Transient Stability Constrained Optimal Power Flow (TSCOPF), or energy function methods. Nonetheless, most of these studies have some limitations, and there is currently no investigation on a system integrated with renewable energy.

Therefore, this requires a new methodology which can properly take into account the uncertainty of the renewable energy power output and transient and voltage stabilities in order to ensure the system security. The objective of this dissertation is to develop a comprehensive scheme for the TTC evaluation by means of the probabilistic method suitable to address all of the above mentioned issues. Hence, one of the main contributions of this research is to fulfill the need and provide system operators and engineers with a suitable analytical tool. The developed method employs Monte Carlo simulation which is justified over other probabilistic methods due to its flexibility in handling and incorporating the uncertainty associated with various kinds of parameters. The TTC is selected based on the risk concept in which the stochastic nature of a system and fault is fully reflected. Voltage stability is ensured by specifying a sufficient margin away from the voltage collapse point, giving a system a room for unexpected incidences. Transient stability is assessed by the time-domain simulation where the detailed models of a synchronous generator together with its controllers, e.g. an exciter and turbine governor, wind power and PV systems can be included. By using the proposed TTC evaluation method, several interesting points regarding the impact of the uncertainty, penetration level of renewable energy, voltage and transient stabilities on the TTC are investigated and discussed.

In addition to the development of the TTC evaluation method, this dissertation proposes a novel efficient method for the TTC evaluation. As is known, the main disadvantage of the Monte Carlo method is its high computation cost as it usually requires a sufficiently large sample size for its convergence. Hence, this time-consuming process is the main impediment to its widespread use for the TTC evaluation. Moreover, the addition of voltage and transient stability assessment results in an extensive computation burden. To overcome such drawbacks, this dissertation proposes several techniques to help speed up the computation, i.e. voltage stability index, system case partitioning using two filters, and decision tree classification.

Commonly used to estimate the proximity away from the voltage collapse point, a voltage stability index is employed in the TTC evaluation to screen out the cases within the Monte Carlo sample set, which are prone to have an insufficient voltage stability margin. With a proper selection of a threshold, a significant number of cases to be checked during Monte Carlo simulation can be reduced. Furthermore, the run time can be saved by reducing the number of cases to be evaluated during Monte Carlo simulation. This is accomplished by system case partitioning. Generally, the system cases in the Monte Carlo sample set can be classified into non-risk-related and risk-related cases respectively, where only the latter are needed to obtain the risk-based TTC value. The system case partitioning uses two filters to grasp the risk-related cases from the Monte Carlo sample set. The first filter is based on the performance indices which are used to measure the severity degree of the system condition and based on which the cases are ranked. The second filter employs a decision tree with one input attribute to roughly screen out the cases prone to transient instability. The results from the two filters are later used to build the partitioned set. Besides, another decision tree with more input attributes is used for fast transient stability prediction. Instead of a time-consuming time-domain simulation, the system stability can be quickly assessed by the decision tree based on the classification patterns it learns from the training data. The performances of the proposed efficient TTC evaluation method in terms of both accuracy and computational speed enhancement are validated and examined via the numerical simulation. The simulation results show that the proposed efficient TTC evaluation method can significantly reduce the run time from the original of 4373 sec to 1255 sec, i.e. saving as much as 71%, while still obtaining an accurate TTC with the error less than 1%.

Finally, the impact of Low Voltage Ride-Through (LVRT) on the transient stability and TTC has also been studied. A generating unit which cannot ride through the excursion of the voltage during the fault is disconnected mainly to protect its equipment from possible damages due to a high fault current, especially power electronics (i.e. an inverter of a PV system). Disconnection of a generating unit is found to affect the system stability; however, depending on several factors. Some faults which cause a drastic drop in the voltage and lead to the disconnection of large generating units, may have a very severe impact on the system stability and frequency stability as the system needs to find generation to make up with the portion lost. This can further lead to a cascading failure phenomenon; hence, an undesirable large-scale power interruption. However, such severe faults may not occur frequently or do not occur at all within the specified lead-time; hence, the impact on the TTC may not be so apparent.

In the future, transfer capability evaluation is still expected to be one of the challenging and important tasks for the system operators as it has always been to maintain the system security and, at the same time, achieve efficient operation. New problems are also expected as a power system evolves with the introduction of new constraints and emerging advanced technologies, renewable energy and communication technologies. The proposed method together with the research findings presented in this dissertation provides key useful information on the TTC which can be later used for achieving efficient operation and planning of a power system. In addition, the proposed TTC evaluation scheme and several speed enhancement techniques can serve as a basis and efficient analytical tool for the future research and development.

本論文は「A Novel Efficient Method for Total Transfer Capability Evaluation of a Power System Integrated with Renewable Energy(再生可能エネルギーを大量導入した電力系統における効率的な送電可能容量算出手法に関する研究)」と題し、6章よりなる。

第1章は「Introduction(序論)」で、本研究の対象である電力系統の送電可能容量(TTC)の一般的概念を説明し、このTTC算出手法について先行研究をまとめている。そして、本論文の目的、構成について述べている。

第2章は「Probabilistic Risk-Based TTC Evaluation(確率的リスクに基づくTTC評価)」と題し、風力発電や太陽光発電などの再生可能エネルギー発電を含む電力系統の運転状態の不確実性とともに電圧安定度制約および過渡安定度制約を考慮した確率的リスクに基づくTTCの算出手法を提案している。ここでは、さまざまな不確実性に基づいた系統状態は、確率分布を考慮したモンテカルロシミュレーションによって生成されている。電圧安定度制約では、まず簡易電圧安定度指標を計算し、不安定と判定された場合は、電圧安定潮流限界点までの潮流マージンを詳細計算し、このマージンが設定した値を確保するまで送電電力を小さくしながら繰り返し詳細計算を行うことにしている。また、過渡安定度制約では、数値積分手法を用いて安定度判別を行い、不安定と判定された場合は、安定と判定されるまで送電電力を小さくしながら、繰り返し数値積分計算を行うものとしている。最後に、再生可能エネルギー電源の大量導入時の一斉脱落を防止するために、これらの電源に備えるLVRT(Low Voltage Ride-Through)機能について述べている。

第3章は「Renewable Energy Modeling(再生可能エネルギー電源のモデル化)」と題し、TTC算出計算において必要となる風力発電と太陽光発電のモデル化について述べている。

第4章は「Efficient TTC Evaluation Method(効率的TTC評価方法)」と題し、本提案手法を実系統に適用するために必要なTTC計算時間を短縮するSystem Case Partitioning手法と決定木手法を提案している。このPartitioning手法では、モンテカルロシミュレーションで扱われる数多くのケースの中から厳しい事故ケースを選別することを、端子電圧、発電機の無効電力出力、送電線の熱容量に関する3つの静的な指標を用いて行い、その絞られた数のケースに対してTTCの詳細計算を行っている。次に、過渡安定度の安定・不安定を数値積分手法で判定するケース数を、CARTに基づいたアルゴリズムを用いて構成した決定木を用いて安定判別を行うことによって大幅に削減する手法を提案している。

第5章は「Numerical Examples and Results(数値例とその結果)」と題し、風力発電や太陽光発電が導入されたモデル系統に対して提案した確率的リスクに基づく効率的なTTC算出手法を適用し、再生可能エネルギー電源出力の不確実性や電圧安定度制約、過渡安定度制約がどのようにTTCに影響するかを数値シミュレーションで明らかにしている。そのシミュレーション結果から、負荷需要や再生可能エネルギー電源出力の確率的変動を考慮すると確率的リスクに基づくTTCは減少すること、再生可能エネルギーの導入率を増加していくと、TTCも増加すること、電圧安定度や過渡安定度の制約を考慮するとTTCが減少することなどが明らかになっている。また、Partitioning手法や決定木手法によって大幅に削減された詳細解析をすべきケースの数とTTC誤差および計算短縮時間との関係も示している。LVRT機能については、TTCに与える影響はあまり見られないが、電力系統の様々なパラメータに依存することを示している。

第6章は「Conclusions(結論)」で、各章の結論をまとめている。

以上を要するに、再生可能エネルギー電源の発電出力の不確実性と電圧安定度、過渡安定度の制約を考慮した確率的リスクに基づく送電可能容量(TTC)を効率的に計算する手法を提案し、再生可能エネルギー電源導入率の異なるモデル系統に対して適用し、TTCに与える影響および計算負荷が大幅に削減されることを明らかにしたもので、電気工学、特に電力システム工学に貢献するところが少なくない。

よって本論文は博士(工学)の学位請求論文として合格と認められる。