Jeremy Lin

Wind Power and Electricity Prices at the PJM Market

Increased deployment of wind-power generation is changing the landscape of power supply around the world. When integrated with electricity markets, wind power is also widely known to influence the prices cleared at the market. This paper examines the effect of wind-power generation on the day-ahead market prices in the PJM electricity market using robust econometric models and statistical inference. Results show that the quantified expected benefits to wholesale market participants may be substantial despite the relatively low wind-power penetration levels still observed within the PJM jurisdiction relative to other markets. The quantified unitary benefits outweigh, by a great margin, the renewable energy credits given to qualifying wind farms across the market. When benefits outweigh the costs incurred to generate them, welfare is accrued. To this end, benefit allocation mechanisms are proposed, which may contribute to the continued development of wind power on the roadmap to a cleaner power industry of the future.

A comparative study of different machine learning methods for electricity prices forecasting of an electricity market

Generally, it is difficult to accurately forecast electricity prices because they are unpredictable. Yet, accurate price forecasting is expected to provide crucial information, needed by power producers and consumers to bid strategically, thereby decreasing their risks and increasing their profits in the electricity market. In this paper, two models using artificial neural networks (ANN) and support vector machines (SVM) were developed for electricity price forecasting. In addition, ant colony optimization (ACO) was used to reduce the feature space and give the best attribute subset for ANN model. Using ACO for feature selection significantly reduced the training time for ANN-based electricity price forecasting model while the results were almost as accurate as those from ANN model.

Potential Impact of Solar Energy Penetration on PJM Electricity Market

Renewable energy resources, such as wind and solar, have become important parts of todays generating resource mix. Along with them, these resources also bring their inherent characteristics of variability and uncertainty. Hence, resource planners must be adaptable to accommodate these resources in conducting resource planning. Understanding their impact on the electric grid must be prior and critical. In this paper, we have studied the potential impact of solar energy penetration in a particular location of PJM electricity market. Key economic and system variables that were analyzed include system production cost, market prices (average zonal prices), congestion (both hours and dollars), and transmission losses. We find that significant penetration of solar energy resources can produce system-wide economic benefits in the respective transmission areas as well as the entire system.

Multiple Period Ramping Processes in Day-Ahead Electricity Markets

This paper proposes an approach to formulate the multiple-period ramping capability of dispatchable generation resources and evaluates the impact of this service on the generation scheduling in day-ahead electricity market. It is discussed that the multiple-period ramping enhances the load following capability of dispatchable generation resources and improves the dispatchability of renewable energy resources in power systems. The presented approach encompasses the uncertainties in the operation scheduling of power systems, using scenario based stochastic security-constrained unit commitment. The presented case studies also highlight the merits of integrating energy storage facilities to reduce the ramping services provided by dispatchable generation resources with respective costs.

LMP Revisited: A Linear Model for the Loss-Embedded LMP

In this paper, the concept of locational marginal price (LMP) is revisited by analyzing the Karush–Kuhn–Tucker (KKT) condition of a general optimal power flow (OPF) problem. The impact of losses on LMP is investigated. Several well-known properties and formulations of the LMP are illustrated from a novel perspective. In particular, the special case of the lossless DC OPF model is discussed. Based on this theoretical analysis, a linear model for the loss-embedded LMP is proposed. The influence of losses on the LMP is recovered by solving a system of linear equations. The proposed method does not rely on strong subjective assumptions, such as the selection of reference buses or the determination of the “loss factor.” Case studies show that the proposed method has distinct advantages compared with the method currently practiced in major electricity markets.

Modeling and simulation of PJM and northeastern RTOs for interregional planning

PJM operates the largest transmission and market system in the U.S. It also borders with other regional transmission organizations, such as Midwest ISO, and New York ISO (NYISO). The issue of transmission planning is not just confined to each RTO internal region, but also crosses borders with neighboring entities. Due to a recent push by federal regulatory authority via Order 1000, it becomes necessary to consider transmission planning (upgrades or new transmission facilities) that touches more than one regional entity. In light of that, it is important to develop model, database and simulation capabilities which can be used to represent and simulate for more than one RTO. One such exercise for PJM is to model three RTOs - PJM, NYISO, and ISO New England (ISO-NE), in a combined and consistent fashion so that we can evaluate the economic merits of any proposed cross-border transmission projects or increase of tie capacities. In this paper, we will show such a model, including preliminary simulation results of base case and other sensitivity cases.

Issues and challenges in smart-grid market operation and simulation

Smart-Grid has been a buzzword in the power industry for some time. While smart-grid could mean differently to different entities, our understanding of it has been quite limited. In this paper, we describe the identifiable developments that seem to relate to so-called smart-grid and that can have impact on the grid and market operations. Obviously, not all developments around smart-grid would have meaningful impact on the market operation, while those that do have impact, have varying degrees of impact, depending on many factors. These various developments include, but not limited to, substation automation, phasor measurement, smart appliance, demand response, electric vehicles (P(H)EV), energy storage, micro-grid etc. We, then, categorize them based on their impact on specific part of power system, such as transmission and distribution, generation and load. We further select those developments that can have significant impact on the electricity market operation as well as market simulation. We then try to present various issues and challenges associated with these selected developments in terms of incorporating them in the market operation and simulation, to account accurately for their unique physical characteristics and economic nature.

Mutually orthogonal LMP decomposition: Analysis of PJM network by null space approach

Under standard approach, locational marginal prices (LMPs) are generally decomposed into system marginal (energy) price (SMP), marginal congestion cost (MCC), and marginal loss cost (MLC). However, earlier work has shown that it is not possible to separate the standard decompositions of LMPs into three mutually orthogonal subspaces. It is also found that MLC term of standard LMP inherently depends on the selection of reference bus, due to their association with only a single direction (one dimensional space) in the set of admissible LMPs. On the other hand, the MCC term, constructed by the null space approach, is found to be independent of choice of reference bus. A key goal of this research is to develop efficient algorithms that provide conceptual and computational simplicity in characterizing the admissible set of LMPs. The key outcome of this work is an efficient algorithm that only requires the identification of a single null vector in a matrix known to be column-rank deficient by one, and hence achieves the goal of low cost computation. In this paper, we illustrate the usefulness of this fundamental null space approach by applying it on well-known PJM network by invoking one constraint at a time. The objective is to analyze the admissible LMP pattern in PJM network, by observing the behavior and relationship of MCC term - both magnitude and direction - to combined SMP/MLC term (marginal energy/loss basis vector).

A Framework for Expansion Planning of Data Centers in Electricity and Data Networks under Uncertainty

This paper presents the expansion planning for data centers and data routes in the data and electricity networks considering the uncertainties in the planning horizon to ensure an acceptable rate of service to the requests received from the end-users in the data network. The objective is to determine the location and capacity of the data centers as well as the required data routes while considering the imposed constraints in the electricity and data networks. The installation cost of data centers and data routes, as well as the expected operation cost of the data centers, are minimized. The proposed problem addressed the uncertainties in the expansion planning of the electricity networks including the availability of renewable generation resources, the variations in electricity demand, the availability of generation and transmission components in the electricity network, and the uncertainties in the number of requests received by the user groups in the data network. The problem is formulated as a mixed integer linear programming problem and Bender decomposition and electricity price signals are used to capture the interaction among the data and electricity networks. The presented case study shows the effectiveness of the proposed approach.

Towards a voltage regulation market

In this work, we propose a possible voltage support or voltage regulation market in which voltage service providers (VSPs) will compete for an opportunity to provide voltage support or regulation by using voltage support devices. We assume that voltage regulation market is a separate but complementary market to existing energy markets (day-ahead and real time markets). An independent system operator (ISO) will decide when the voltage regulation market is to be launched. ISO will determine the required voltages (in p.u) at some nodes and clear the market based on the submitted bids and offers by the voltage service providers. ISO can anticipate the low-voltage situation ex-ante at some locations and can initiate the voltage regulation market. The voltage regulation market problem was formulated as nonlinear AC-OPF model and was solved using improved differential evolutionary algorithm. We used an IEEE 14 bus system as a test system in which the voltage regulation market clears. We showed that the development of voltage regulation market is of practically possible and can improve both social welfare and system condition.