abstract: This thesis pursues a method to deregulate the electric distribution system and provide support to distributed renewable generation. A locational marginal price is used to determine prices across a distribution network in real-time. The real-time pricing may provide benefits such as a reduced electricity bill, decreased peak demand, and lower emissions. This distribution locational marginal price (D-LMP) determines the cost of electricity at each node in the electrical network. The D-LMP is comprised of the cost of energy, cost of losses, and a renewable energy premium. The renewable premium is an adjustable function to compensate `green' distributed generation. A D-LMP is derived and formulated from the PJM model, as well as several alternative formulations. The logistics and infrastructure an implementation is briefly discussed. This study also takes advantage of the D-LMP real-time pricing to implement distributed storage technology. A storage schedule optimization is developed using linear programming. Day-ahead LMPs and historical load data are used to determine a predictive optimization. A test bed is created to represent a practical electric distribution system. Historical load, solar, and LMP data are used in the test bed to create a realistic environment. A power flow and tabulation of the D-LMPs was conducted for twelve test cases. The test cases included various penetrations of solar photovoltaics (PV), system networking, and the inclusion of storage technology. Tables of the D-LMPs and network voltages are presented in this work. The final costs are summed and the basic economics are examined. The use of a D-LMP can lower costs across a system when advanced technologies are used. Storage improves system costs, decreases losses, improves system load factor, and bolsters voltage. Solar energy provides many of these same attributes at lower penetrations, but high penetrations have a detrimental effect on the system. System networking also increases these positive effects. The D-LMP has a positive impact on residential customer cost, while greatly increasing the costs for the industrial sector. The D-LMP appears to have many positive impacts on the distribution system but proper cost allocation needs further development. / Dissertation/Thesis / M.S. Electrical Engineering 2011

An approach to utilize Distributed Generation (DG) to minimize the total load shedding by analyzing the power system in Transactive energy framework is proposed. An algorithm to optimize power system in forward and spot markets to maximize an electric utility’s profit by optimizing purchase of power from DG is developed. The proposed algorithm is a multi-objective optimization with the main objective to maximize a utility’s profit by minimizing overall cost of production, load shedding, and purchase of power from distributed generators. This work also proposes a method to price power in forward and spot markets using existing LMP techniques. Transactive accounting has been performed to quantify the consumer payments in both markets. The algorithm is tested in two test systems; a 6-bus system and modified IEEE 14-bus system. The results show that by investing in DG, utility benefits from profit increase, load shedding reduction, and transmission line loading improvement.

Cette thèse vise à développer les opportunités et conséquences d’une demande incertaine pour le réseau de transport de gaz. Ce sujet est décliné en quatre contributions. Les deux premières adoptent une perspective de long terme : on cherche à évaluer l’efficacité de la réglementation du taux rendement lorsqu’il s’agit d’inciter à la réalisation de projets d’infrastructures gazières dans des pays en développement. Une première contribution analytique présente le développement d’une représentation simplifiée du réseau de transport de gaz, de forme Cobb-Douglas. Inspiré par les projets d’acheminement de gaz naturel au Mozambique, celle-ci est ensuite utilisée pour évaluer dans quelles conditions il est possible pour une autorité de régulation de choisir un taux de rendement régulé qui améliore l’efficacité du système dans le cas où la demande réelle serait plus importante que la demande anticipée par la firme régulée. A moyen terme ensuite, l’efficacité face à une demande de plus en plus variable de la structure tarifaire actuelle dite « entrée-sortie » pour l’accès au réseau européen est évaluée. Après avoir démontré l’existence d’inefficacités dans un tel système, celles-ci sont évaluées numériquement. Enfin, la dernière contribution explore la possibilité d’offrir directement la flexibilité du réseau de transport de gaz à ses utilisateurs, dans le cadre d’enchères et du système de prix nodaux. Après avoir souligné la complexité d’un tel mécanisme, les limites à son efficacité sont présentées. A chaque fois, l’analyse repose sur la modélisation simultanée du réseau de transport de gaz (en régime statique ou transitoire) et des mécanismes économiques en jeu. / This PhD thesis is centered on the opportunities and impact of demand uncertainty for the gas transport networks. We study the ability of various market designs to foster an efficient network allocation in liberalized gas markets when demand is variable or uncertain. We introduce and solve operation research models that bind an economic representation of the gas market and its associated regulation, to a technical representation of the gas network. The complex interactions at stake in liberalized gas markets, where shippers trade gas for its economic value and coordinate with system operators that allocate and operate the network, result in MCP or MPEC formulations. While a detailed network representation is necessary to assess the feasibility of gas flows under any market organization, the physics and engineering of gas transport networks adds non-linearities and non-convexities to those already challenging formulations. This thesis is divided in four contributions. We first introduce an approximated network representation of the Cobb-Douglas form and use it to study the impact of long-term demand uncertainty on investment problems in developing markets subject to rate-of-return regulation. We then study the effect of demand variability on daily gas dispatch in the European Entry-Exit system, using a linearized steady-state network representation. Finally, we assess the benefits of introducing flexibility products in gas locational marginal pricing auctions to handle intraday demand uncertainty. This requires the use of a linearized transient network formulation to account for linepack dynamics.