Slides for PSERC Webinar on Jan. 21, 2014

In power system operations, the industry has long employed the criterion that the system, following any “ credible” contingency, must satisfy certain performance criteria, e.g., any “ N-1” contingency must result in flows on all circuits within their emergency ratings. This criterion is employed within the security-constrained economic dispatch (SCED), which is the computational engine for real-time electricity markets. For a given unit commitment and a given loading condition, the SCED is thought to achieve the most economic conditions that also provide a satisfactory level of security, implying that better economy may be achieved only through less security and that more security may be achieved only through less economy. In this seminar, we show that this is not so; i.e., there exist operating conditions that are both more economic and more secure than the operating conditions determined by the SCED. In doing so, we present the risk-based security-constrained economic dispatch (RB-SCED), an optimization formulation that extends the SCED by enabling adjustment of emergency limits on individual circuits while simultaneously imposing a constraint on system risk. The more economic and more secure operating conditions from RB-SCED are shown to out-perform operating conditions from SCED in terms of cascading overloads, post-contingency angular separation across lines, and voltage instability margin. We also illustrate that locational marginal prices from RB-SCED contain an additional component which more appropriately incentivizes market behavior. Replacement of SCED by RB-SCED in today’ s real-time markets would result in reduced energy cost and improved reliability performance.

This research is one of the PSERC projects coordinated by the Consortium for Electric Reliability Technology Solutions (CERTS) with funding provided by the U.S. DOE.

Slides for PSERC Webinar on Feb. 4, 2014

This talk addresses attribute preserving power network equivalents, with a specific focus on the relatively unexplored area of transmission line limit preserving equivalents (LPEs). There is a desire to develop LPEs for power system interconnections to be used in markets and reliability studies. Present day equivalencing algorithm, such as Kron reduction, do not retain the limits for the transmission lines as they are equivalenced. This talk explores techniques to determine limits for the equivalence lines.

This research is one of the PSERC projects coordinated by the Consortium for Electric Reliability Technology Solutions (CERTS) with funding provided by the U.S. DOE.

PSERC Webinar on Feb. 18, 2014

Cascading outages have resulted in significant disruptions to power systems all over the world. Controlled islanding can provide a corrective measure of last resort to alleviate the impact of large disturbances. This project presentation provides detailed demonstrations of using slow coherency based controlled islanding to prevent cascading outages in bulk power systems. An integrated algorithm was developed to identify cutsets for large scale power systems for the application of slow coherency based controlled islanding schemes. The large scale power system was represented as a graph and a simplification algorithm was used to reduce the complexity of the system. Generators belonging to the same slowly coherent group were collapsed into a dummy node, and a graph partition library was used to split the graph into a given number of parts. Some extra islands formed by the partition library were merged into their adjacent large islands and the original cutset of the actual power system was recovered from the highly simplified graph. A software package was developed based on the algorithm, and dynamic simulations were run on the Western Electricity Coordinating Council (WECC) system to verify the effectiveness of the cutsets obtained. Finally, to test the islanding performance, four extreme contingencies under two different operating conditions of the WECC system were tested using the time domain simulations. The cutsets used in the controlled islanding cases were obtained from the software package developed using the graph partition library. The time domain simulation results for the four contingencies with controlled islanding and uncontrolled islanding will be shown, and the dynamic performance in each case analyzed. Further analyses were conducted to examine the amount of load shed in each case. A discussion is provided of the cutset sensitivity and time sensitivity of islanding along with practical implementation issues and conclusions.

Slides for the PSERC Webinar on March 4, 2014

The consequences of environmental and energy policies in the U.S. can be severely constrained by physical limits of the electric power grid. Because of the complexity of power grid operation, computing limitations have until very recently made it impossible to solve a policy analysis or planning model that combines realistic modeling of flows with a detailed transmission system model and the prediction of generator investment and retirement. We construct and solve a model of the eastern US and Canada that combines these characteristics. Then, because a smaller model might be for some additional purposes, we explore the effects of transmission model simplification on the accuracy of simulation results. To evaluate the amount of detail necessary, we simulate the short- and long-term effects of imposing a price on the carbon dioxide emissions from the power plants in nine northeastern US states, as the Regional Greenhouse Gas Initiative does. We consider three grid models that simplify the actual 62,000-node system to varying degrees. Our 5000-node model matches the 62,000-node model very closely. We use it as the basis for evaluating the more simplified models: a 300-node model and a model with just one node, i.e. no transmission constraints. With each of the three models, we predict the carbon dioxide emission impacts, electricity price impacts, and generator entry and exit impacts of the emission price, over the next 20 years. We find that most of the impact predictions produced by the 300- and one-node models differ from those of the 5000-node model by more than 20%, and some by much more.

Slides for the PSERC Webinar on April 1, 2014

Pressing environmental issues and concerns over energy security are driving worldwide interest in renewable energy resources. Unlike conventional generation, however, power from wind and solar resources is inherently variable in its supply. It is non-dispatchable, highly intermittent, and difficult to forecast. This intrinsic variability in supply represents the most important obstacle to the large-scale grid integration of renewables.The first part of this webinar will provide a system theoretic perspective of the critical challenges facing deep renewable energy integration and formally examine the role of energy storage in mitigating the attending costs of integration. In particular, we aim to quantify the value of energy storage capacity within existing market constructs. The second part of the webinar presents the results of an application of the Cornell SuperOPF to evaluating the system benefits of distributed storage for mitigating the uncertainty of wind generation. 

PSERC Webinar Slides, April 22, 2014

Aggressive development of renewable electricity sources will require significant expansions in transmission infrastructure. But how many new circuits, and where and when? These decisions must be made in the face of profound economic, technological, and political uncertainties. I present a stochastic two-stage optimization model that captures the multistage nature of transmission planning under uncertainty and use it to evaluate interregional grid reinforcements in Great Britain and the U.S. Western Electricity Coordinating Council (WECC).  This work was supported in part by the Consortium for Electric Reliability Technology Solutions (CERTS) with funding provided by the U.S. DOE.

PSERC Webinar Slides, May 6, 2014

With the ongoing investments in a smarter electric grid, new algorithms and devices are being developed. Synchrophasor applications for electric transmission systems are one of the most critical smart grid technologies. Synchrophasor devices provide synchronized measurements at high rates for enhanced wide area situational awareness and decision support using new applications. High quality of synchrophasor measurements is vital for most of these applications, especially for real time control. To test the performance of synchrophasor devices and applications, real time modeling and simulation is required. This talk will present research activities related to building a real-time wide area monitoring and control test bed, and using it for testing and validation of synchrophasor devices and voltage stability monitoring applications. Additionally, this talk will discuss development of two new tools using synchrophasor data: the Phasor Measurement Unit Performance Analyzer (PPA) and Real Time Voltage Stability Monitoring (RT-VSM).

Slides for PSERC Webinar on Sep. 16, 2014.

A new approach to protection will be presented, one that will lead to simplified but secure and reliable protection schemes by fully utilizing existing and expected technology advancements. The dynamic state estimation based approach provides a secure and dependable protection scheme. It does not require coordination with other devices or protection schemes, and settings are minimal (comparable to differential protection); thus the term “ setting-less” . The dynamic state estimation based approach requires complex analytics to be performed on the data acquired with the data acquisition system of the relay. The research has also proved through numerical experiments and laboratory experimentation that the analytics can be performed within the sampling period of typical data acquisition systems for relays with substantial margin. The technology promises to drastically reduce relay mis-operations, solve protection gaps and in general increase the reliability of power systems. We will discuss status of this technology, planned field demonstrations and the overall program direction. 

Slides for the PSERC Webinar on October 7, 2014

Wide-spread deployment of synchronized phasor measurements at substations in the past decade facilitated their utilization for various network applications one of which is the state estimator. Early investigations looked at cases where synchronized measurements were used to augment the existing SCADA measurements, while more recent studies consider the scenario where sufficient number of strategically placed phasor measurements enable state estimation based solely on phasor measurements. In this talk, we will first review state estimation formulation that uses only phasor measurements and present a solution algorithm that remains robust against missing or erroneous phasor measurements. We will then present some of the benefits of having phasor measurements (as opposed to SCADA measurements) for parameter error identification. We will conclude the talk by addressing the issue of non-detectable and/or non-identifiable random or malicious errors in phasor measurements and illustrate some simple solutions to avoid manipulation of flow estimates on network branches via injection of such errors in phasor measurements.

Slides for the PSERC Webinar on October 21, 2014

The energy efficiency of houses in a demand response program can affect consumer comfort and the effectiveness of the demand response program. Houses with three levels of thermal integrity were modeled with both properly-sized and oversized air conditioners. Improved thermal integrity significantly reduces both peak demand and energy consumption of all the houses modeled. A demand response program that cycled air conditioners off for varying durations during the summer cooling season was simulated for the various house designs. Without demand response cycling, houses with oversized air conditioners maintained occupant comfort, with the possible exception of humidity, which was not simulated, throughout the summer. When air conditioners were cycled, occupant comfort was degraded, with houses having improved thermal integrity providing greater comfort. If air conditioners were properly-sized for the house design and thermal integrity, occupant comfort on the hottest days was degraded regardless of demand response cycling. Occupant comfort was worse, however, with cycling. For these houses, better thermal integrity slightly improves occupant comfort without demand response, and significantly improves it with demand response cycling. These results should be considered in the design of demand response programs, and the simulations should be expanded to include other days, locations, and house designs.

Slides for the PSERC Webinar on November 4, 2014

Declining prices coupled with tax incentives are driving increased use of distributed energy resources in the electrical utility system. As a result, more power is being generated in homes, businesses, and commercial buildings and used locally. This jump in power production at the distribution level presents a challenge to the traditional electrical transmission and distribution system based on centralized power generation and control. The centralized system operated by utilities was not designed for the flexible load tracking required by renewables or the control of large numbers of distributed electrical energy resources. An alternative approach that holds promise is a dynamic distribution system that includes a Distribution System Operator as the local balancing authority. As conceived, the dynamic distribution system uses local sources to track loads, stabilize voltage and frequency, and smooth intermittent renewable energy generation providing a predictable, constant load profile to the utility. This new dynamic distribution system connects central and local electricity generation, storage, microgrids, and loads with a marketplace that enables energy transactions, such as payments passing between buyers and sellers of energy at the local distribution level. This new system provides a promising framework for distributed energy resources to deliver the same services at a better price, with decreased power losses, decreased emissions, and better reliability. The challenges and opportunities associated with this dynamic distribution system architecture will be discussed in this presentation.

Slides for the PSERC Webinar on November 18, 2014

High penetrations of photovoltaic (PV) systems impact distribution system operations in a number of ways, spanning voltage and other power quality, resistive losses, requirements for transformer bank and conductor capacity, reverse power flow and protection equipment operation and transformer aging. This talk begins with a simulation-based investigation of these impacts at varying PV penetrations and across a diversity of distribution system topologies and climates in California. The simulations are driven by a unique distributed PV data set that includes real 1- and 15-minute production from small PV systems across the scale of a distribution feeder. We first interpret the impacts in terms of engineering metrics, including voltage excursions, peak capacity requirements, and resistive losses. Then, using cost and circuit-level peak load growth data from Pacific Gas and Electric, we interpret the impacts in economic terms. We find that in most cases the economic impacts are small relative to avoided energy costs. However, we do see that in a limited number of cases the voltage quality impacts are moderate. Using this result and other anecdotal evidence as motivation, in the second part of the talk we discuss a new approach for volt-VAR optimization with reactive power capabilities of PV inverters. Relative to existing approaches, the strategy we describe is decentralized and model-free but identifies optimal setpoints for inverter reactive power across a variety of objective functions. This work is joint with Michael Cohen, Paul Kauzmann and Daniel Arnold.