The Carnegie Mellon Electricity Industry Center sponsors a weekly seminar series on Thursdays at 12:15 PM in the EPP conference room (Baker Hall 109). These seminars provide an informal means of presenting research, listening to outside speakers, and learning about the electricity industry. Upcoming and past seminar speakers and topics are listed below. When possible the slides for the talk are also available.
If you plan on attending a seminar, please contact Marilyn Walgora (firstname.lastname@example.org), so we can order sufficient food for lunch.
If you would like to give a seminar, or have a good idea for a seminar topic, please contact Jay Apt (email@example.com or 412-268-3003).
"Retail Choice in New Zealand"
"Do We Understand the Risks in Supervisory Control and Data Acquisition (SCADA) Systems?"
Software Engineering Institute
University of Virginia
Data collection, control, communication, and management, all of which are essential for the effective operation of large-scale infrastructures, are increasingly being performed by supervisory control and data acquisition (SCADA) systems. These are complex communications systems, interactive between human and hardware/software components, and highly vulnerable to the threat of cyber-terrorism. Critical infrastructures, such as railways, water, electric power, oil and gas pipes, and telecommunications, are interdependent and interconnected. Thus, a terrorist intrusion into the SCADA system that controls, for example, a railway system, may be able to transmit malicious disinformation that can result in a major disruption and loss of lives and commerce.
This paper, addresses the risk-assessment process associated with SCADA systems. To fully explain the risk-assessment process, the basic configuration of SCADA systems is described, leading to their inherent vulnerabilities associated with commercial off-the-shelf (COTS) hardware components, communications, and human supervisory control, among others. The risk-assessment process is performed with Hierarchical Holographic Modeling (HHM) to systemically identify the myriad sources of risks. Three major sub-HHMs are identified: hardware and software of SCADA systems, human supervisory control associated with SCADA systems, and the environment within which SCADA systems function.
"Structural Reliability of Electric Transmission Lines - Is it a Problem?"
Anthony M. DiGioia, Jr.
GAI Consultants, Inc.
Do electric transmission lines in the United States and around the world experience structural failures? What are the potential causes of failures? Are the risks and associated costs of failures significant? Can the risk and associated costs be quantified? These questions will be discussed at the meeting. Current transmission line structural reliability design criteria will be discussed and examples of structural failures from around the world will be presented. Finally, an open discussion will be held concerning the future role that CMU could play in resolving this issue.
"Is it all about Economic Profitability? Insights on the Diffusion of Building-integrated Photovoltaic Technologies into Urban Areas Following an Innovation-diffusion Theory Approach"
Maria João Rodrigues
Visiting Scholar in the Department of Engineering and Public Policy and Researcher at the Centre for Innovation, Technology and Policy Research, IN+ at IST, Technical University of Lisbon
The introduction of renewable energy technologies into the electricity systems is being fostered by governments worldwide mainly as a response to climate change. In Europe these policies are seen to be more stringent given its commitment to the Kyoto protocol goals. As a consequence ambitious development strategies can be found both at local, regional, national and European levels, while being commonly differentiated by technology. With respect to photovoltaic (PV) solar technologies, the systems focused in this paper, Germany is undoubtedly leading the way in installed capacity, both in absolute and capitation terms. At the end of 2002, Germany attained an impressive 277 MW PV cumulative installed capacity, about 20% of the world installed capacity, and a capitation of 3,37 W per person. Besides Germany, both Switzerland and the Netherlands ought to be referenced in the European context, with capitations of 2,67 and 1,64 W per person respectively. These countries share one common feature in this unquestionable success: the existence of ambitious and concerted public policies targeted at the same PV market segment ? grid-connected distributed applications. Although consisting of very different approaches, namely concerning the dichotomy between public intervention and the marketplace, these policies recognize the outstanding opportunities offered by PV technologies in the built environment. Unlike other electricity-generation technologies, the use of PV as an integral part of a building (commonly referred to as building-integrated PV applications or BiPV) provides unique opportunities for exploiting both energy and non-energy benefits, both believed to add value to the electricity generated.
Despite of the impressive increase in installed capacity of BiPV systems worldwide, mainstream commercialization has not been achieved yet nor has the market potential been substantially realized. In other words, the technology is still in its first stages of the diffusion process.
The discussion on how to mainstream PV technologies into the urban environment is presently at the top of the research agenda in Europe and, to some extent, internationally, in particular at the International Energy Agency (IEA) level. This presentation aims to contribute to the above discussion by providing an interpretation of BiPV diffusion into urban areas in the light of the theory of diffusion of innovations of Rogers (1995). The analysis provides new insights to factors and actors that may accelerate or influence the technology rate of adoption other than the value of the electricity generated.
"Statistical Characteristics of Cascading Failures in Power Networks"
Electric power systems consist of thousands of interacted components, distributed geographically. On the one hand, it is the interactions among components that realize long-distance economic transmission of electrical energy and complete the service to millions of customers. On the other hand, the interactions may introduce unexpected consequences to the interconnected power systems under certain unfavorable circumstances, such as defected structure of the system, extreme weather, abnormal load variations, hidden failures of protection systems. In some cases, it causes collapse of the power systems. In this talk, after briefly reviewing the scenarios of August 2003 Blackout in US, some statistical characteristics of the cascading failures of power systems are investigated in a regular resistor network and a 3357-bus practical power system respectively by extensive computer simulation. To represent the cascading sequences and the relationship among element failures, a graphical representation based on Boolean networks is used, which could be a promising way to depict the vulnerability of power networks under adversary disturbances.
"Electric Cooperatives: An International Success Story"
Executive Director of Research and Technical Services National Rural Electric Cooperative Association
"Creation of US Industry Sector Consumption Mixes: Disagregating Electricity Generation and Estimating Interstate Electricity Transfers"
CMU EPP and CEE
"Blackouts and Electric Reliability - What Should We Learn?"
The August 14, 2003 blackout and the investigation that followed illuminated many problems with current electric system operations and reliability practices. Alison will talk about what caused the blackout, what needs to be done to improve grid reliability in North America, and what it will take for those solutions to be implemented.
"The Operation of Distributed Generation in the Restructured Power Industry"
Small-scale distributed generating technologies are gradually replacing conventional generating technologies for some applications in the electric power system. The prospect of independent ownership for these distributed technologies is being encouraged by the current deregulation of the industry, and it is likely that these new generators will be independently operated as well as independently owned. This presentation investigates locating numerous small-scale generators in distribution feeders and the possible impact of these facilities on the stability of the distribution system. Simulations demonstrate, unexpectedly, that a small load disturbance is capable of causing frequency instability in the primary dynamics of the distributed generators. Eigenanalysis of the instability suggests that it is a system, rather than individual facility, phenomenon. A method to regain system stability along with an example of implementing this method is presented.
With respect to the emerging competitive markets, there is interest in developing both the technologies and market structure necessary to allow distributed generators to participate in energy and ancillary service markets. This presentation introduces the use of a closed loop price signal designed to coordinate generator actions in the competitive market while also maintaining the desired level of system reliability and stability. Price signals are one mechanism available to coordinate the operation of the power system in the emerging competitive market. Results presented will demonstrate the role of price in coordinating both the engineering and the economic aspects of distributed generator operation in a restructured power system. This paper demonstrates the ability of the distributed generators to participate in a future competitive energy market via a price signal coordinating system operation.
"Energy Infrastructure and Security"
Concerns about how to safeguard key infrastructures (energy, communications, banking, roads, etc.) from deliberate attack are longstanding, but over the last decade increasing emphasis has been placed on the possible impacts of terrorism. Activities to address these concerns are sometimes called Critical Infrastructure Protection (CIP), which is somewhat different from the longstanding concept of ‘energy security,’ which focuses on politically- and economically motivated supply interruptions. Energy infrastructures differ somewhat in their principal security concerns. Electricity systems are among the most difficult of infrastructures to safeguard because of the infeasibility of storage, their complexity and the potential for cascading failures. Parts of the oil and gas infrastructures are among the most concentrated in the world, especially production and refining. Nuclear infrastructures pose risks that are uniquely global and dreaded by the public. This review discusses how energy infrastructure and security are related, how it differs from most traditional energy security terms, and what it may mean for private and policy decisions. Key concepts include redundancy, diversity, resilience (or security), storage, decentralization, and interdependence. The concept of CIP is still relatively new and is likely to evolve over time, possibly away from a ‘guards, gates, and guns’ defensive approach and towards a design approach that yields systems that are inherently harder to successfully attack. Such survivable systems may feature distributed intelligence, control, and operations.
"Investment in U.S. Electric Utilities under Regulation and Natural Gas Price Uncertainty: Timing of Plant Retirement and New Technology Choice"
Electric utility managers confront impending replacement of an aging coal generation fleet in a business environment characterized by volatile natural gas prices and uncertain future regulation of greenhouse gases and conventional pollutants. Generation technology choice and the timing of investment are examined with a stochastic dynamic programming model, which minimizes the present value of future power generation costs. Regulatory uncertainty combined with irreversible investment decisions creates an incentive to delay coal plant retirement with associated power generation cost and pollution consequences. The programming model provides a platform to examine these expected welfare loses arising from regulatory uncertainty. Conversely, irreversible investment prior to resolution of regulatory uncertainty can impact the effectiveness of eventual regulations.
"Market Power and Mitigation in Deregulated Electricity Markets"
Carnegie Mellon University, Engineering and Public Policy and Electrical and Computer Engineering
Conventional measures of market structure used by economists, such as the Herfindahl Hirschman Index (HHI), give a misleading picture of the competitiveness of electric power markets, since these metrics do not consider the special properties of electricity as a commodity. This paper uses an alternative metric, based on the interplay between the system capacity as a whole, the capacity of individual suppliers, and the level of demand, to evaluate the competitiveness of three recently-deregulated electric power systems. Since it is possible for sufficiently large suppliers to set prices arbitrarily high by threatening to withhold generation, it is possible to see the exercise of market power even in times of surplus capacity. An analysis of California, PJM, and New York between June 2000 and June 2001 finds that none of them can be regarded as highly competitive, contrary to what conventional measures of market power would suggest. Five candidate market-power mitigation measures are discussed within the context of the California, PJM, and New York electric power systems. All five options will raise electricity costs, though the benefits from deregulation in some cases may outweigh the costs of mitigation. More importantly, different mitigation options will be less costly in different power systems. The likely success rate of each mitigation scheme also varies depending on the properties of the system to which it is applied. For example, building additional transmission to stifle market power through increased imports is likely to be more successful in California, whose neighbors experience noncoincident peak demands, and less successful in New York and PJM. The lessons for regulators is that decisions on deregulation of electric power markets and market-power mitigation should be made on a system-by-system basis, and that applying one set of rules to all systems will increase costs and decrease operating efficiency.
"A Multi-Layered Approach to Transmission Provision and Pricing in the Electric Power Networks"
CEIC, Engineering and Public Policy and Electrical and Computer Engineering
In this talk, I review three qualitatively different mechanisms of delivering electric power under open access. The first approach is based on optimizing power dispatch under transmission constraints, and providing a bundled electricity price signal which incorporates both energy and systems support charges. This approach is based on the original notions of spot electricity prices  underlies today’s spot markets in several parts of the U.S. electric grid and is recommended by Professor Hogan at Harvard [1a]. The second approach allows for the electricity trading process to be separate from the transmission system support needed to deliver the traded power. This was introduced by several Berkeley faculties in . The only constraint is that the market participants trade under the technical constraint that the transmission limits are not exceeded. There is no transmission price signal in this method. Finally, the so-called two-level transmission provision and pricing was introduced by Ilic et al at MIT in . This method is based on iterative information exchange between the market participants and the transmission system provider: The market participants inform a system provider concerning the location and amount of power they wish to inject into particular locations within the electric grid, and the system provider, based on all given requests, optimizes use of the available transmission capacity and sends the transmission price signal to the market participants. The market participants adjust their requests, the delivery price gets adjusted, and the transactions are implemented. It is documented in  that at the equilibrium all three schemes result in the same optimum under several simplifying assumptions.
In this paper we review the assumptions under which these transmission provision and pricing schemes are designed and compared to:
- Analyze current industry proposals for transmission provision and pricing in light of the three methods.
- Propose a generalization of the method described in  which allows for a multi-layered reliability-related risk management and valuation of system support.
- Summarize recent simulation results [4, 5, 6] illustrating typical outcomes of the multi-layered transmission provision and pricing.
"Re-Regulation in California"
The California energy crisis in 2000-2001 and the discovery of the Enron gaming strategies applied in CA showed that the California energy markets are far from efficient. The California ISO is currently working on a new market design (MD02, with its final implementation date delayed until 2005) which is supposed to fix the drawbacks of the old one. I will compare the current market design with the proposed one and examine some implications for California utilities.
"Efficient Pricing of a Bundled Product of both Real and Reactive Power"
West Virginia University Department of Economics
This paper examines price mechanism with one price assigned for each level of bundled real and reactive power. This pricing approach is simulated on the simple 3-bus system power auction where generators provide their bids on the bundle of real and reactive power. System operator (SO) is able to dispatch generators efficiently when the generators bid competitively. Incentives to exercise market power with respect to reactive power are tested on the auction. In addition, 30-bus network was tested, with the purpose of identifying generators willing to raise reactive power bids.
The Washington Post
"Cyberspace Influence on the Aug 14 Blackout"
Software Engineering Institute
Of the many indicators, causes, and participating factors that existed in the US and Canadian electric infrastructures, one of the most neglected is the role Cyberspace plays in the proper operation of the power grid. Many "root causes" have been blamed for the Aug 14 NE blackout, but the true point of no return for this event appears to have been caused by a software fault in the EMS system for one particular company. Although there is no evidence in this case of an external cause for this fault, study of the fault can illustrate how vulnerable the grid is to Cyberspace manipulation. This presentation and discussion will focus on the interaction of the electricity and cyberspace infrastructures.
"The Future of Nuclear Power"
Professor of Physics and Director of Energy Studies, MIT
"Duquesne Light Operations and Business Model"
Senior VP and Chief Operations Officer of Duquesne Light
"The Pittsburgh Supersite Program: Characterization of Aerosols"
Professor of Civil & Environmental Engineering at CMU
In 2000, EPA established seven temporary Supersite Airborne Particle Monitoring Stations around the U.S., one of them in Pittsburgh. Additional funding for the Pittsburgh supersite was provided by DOE for sampling major sources. The monitoring program here operated from July 1, 2001 to September 30, 2002. Besides baseline sampling conducted continuously over the 15-month period, there were two intensive sampling runs during July 2001 and January 2002. More than 30 samplers were used during baseline and intensive sampling to obtain data for various aerosol and gaseous chemical species. Instrumentation included time-integrated monitors, such as filters that sampled for later chemical analysis, as well as continuous and semi-continuous monitors to capture short-term variability in concentrations. Results in a few categories of sampling are presented in this seminar. First, we will explore the relative importance of local sources versus long-range transport influencing PM2.5 concentrations in Pittsburgh (PM2.5 = particulate matter with diameters less than 2.5 micrometers). Then we will compare continuous and time-integrated data, investigating what we can learn from highly time-resolved information. Finally, we will consider the chemical composition of aerosols in Pittsburgh. We will explore what this information can tell us about developing control strategies to reduce PM2.5 concentrations.
"How to Write an Op-Ed Piece"
"Autonomous Agents in Electricity"
Sarosh Talukdar and Paul Hines
Systems of autonomous agents (also called distributed decision-making systems, and complex adaptive systems) have many natural and the artificial manifestations. Immune systems, insect societies, artificial life, electric grids and economies are some examples. In this talk, Sarosh will explain some of the open questions in designing systems of autonomous agents, particularly, questions about emergent behaviors, phase-transitions, task-decomposition, and cooperation. Paul will describe
his progress in answering the questions on task-decomposition and cooperation.
The Washington Post
"An Examination of Capacity Markets in Electricity"
Alex Galatic and Frank Lacey
Demand for electricity constantly fluctuates and depends a great degree on one unpredictable factor - the weather. Tomorrow's weather might be relatively certain, but next summer's weather is less so. At the same time, supply is vulnerable to equipment failure that may, at least temporarily, remove a significant source of electricity from the market. As millions of people experienced the blackout on August 14th, 2003, electricity is an important, if not essential, commodity.
Uncertain demand coupled with uncertain production capability for a valuable commodity indicates a need for storage, but electricity is consumed at the same instant that it is generated, and it cannot be stored in significant quantities. In other words, electricity supply cannot be inventoried - at least not directly. Generating facilities fill the role of inventory, so, although electricity in its final form is not stored, in a way, it is stored as fuel in coal piles, natural-gas storage fields, and oil tanks, standing ready to be converted into electric energy.
We depend on the idle production capacity of generators to provide the inventory necessary to compensate for the loss of a large production facility or an abnormal increase in demand due to extreme weather - and keep the lights on.
Some regions in the Northeast have attempted to create "capacity markets" intended to encourage the development of supply reserves adequate to protect reliability, but they have been ineffective at best, and perhaps even create disincentives to build reserve capacity. This discussion examines why capacity markets in the Northeast fail to encourage the development of new generating capacity, and even worse, how they harm competition.
"A Comparison of the Ongoing Restructuring of the Domestic Natural Gas and Electricity Industries"
Energy Center Pittsburgh
The premise of this presentation is that the federal and state efforts to restructure the electricity industry haven't adequately taken into account the different 20-year history of federal and state efforts to restructure the natural gas industry; and, yet, the current failures to create viable competitive retail markets in both industries have common roots or causes.
The gas industry restructuring or deregulation process commenced in the late 1970's as a result of gas shortages during that decade. The gradual deregulation of gas wellhead prices occurred as the result of federal legislation. During the 1980's, as the industry reacted to unexpected surpluses of gas and lower than anticipated prices, various pieces of federal and state regulation provided for an evolution to wholesale and retail markets, the infrastructures of which took years to put in place. This process was aided by the fact that the retail market being created was for industrial and large commercial customers only (until the mid 1990's).
Though the antecedents for the power industry restructuring do go back to 1970's federal legislation and a Supreme Court decision, it was the action of various state legislatures in the mid 1990's that really made electricity restructuring take off. Legislators were convinced that lower prices could be experienced for all retail customers - residential, commercial, and industrial alike - instantaneously, with the passage of legislation and the writing of enabling regulations. The proper development of wholesale markets, let alone the necessary infrastructure, was not considered.
At this point, both gas and power competitive retail markets in just about every state are almost non-existent. If any market is functioning at all, it is doing so marginally. The reasons for this set of circumstances are basically twofold. First, state regulators were unable to change their historic command and control regulation model to one that would allow markets to function and thereby protect consumers better than could regulation. Secondly, the growth of competitive retail markets floundered on the inability of federal regulation to complete the creation of wholesale markets. In gas, though the commodity wholesale market was reasonably well established, the capacity wholesale market never came into being. In power, the evolution from loose power pools to tight power pools, to ISOs, to rational RTOs with well-functioning capacity markets, though still proceeding in some parts of the country, may never develop as federal regulators intended into a Standard Market Design.
Certainly, there are other reasons for the loss of the power and gas restructuring or "deregulation" momentum, such as the proliferation of price caps, the actions of incumbent gas and power utilities, the meltdown of wholesale trading in both gas and power, and the accusation of improper or illegal activities by certain entities. But, to this presenter, the two mentioned above are the more basic causes of the current situation.
"Information Technology and Power Distribution/Consumption"
In this talk, I will present on the potential and role of information technology (IT) for the power sector. Specifically, I will examine the interaction of IT with power distribution and consumption -- setting aside issues of IT usage at the pool (transmission) level. I present a brief overview of the technology, desired services, and current status, and highlight some issues. Beyond automatic meter reading (AMR), I consider IT capabilities for control, operations, and new services. Extending real-time control to the appliance level might have dramatic impact on power system stability and costs. According to one estimate, reducing the peak load by a few percent can reduce the costs of electricity by over 20%.
While many of the new technologies are gaining commercialization, integrated solutions are not widespread. In this environment, I present a preliminary analysis of the potential of such technologies for developing countries. Here, given they often lack what is traditional equipment in the West (like automatic reclosers, capacitor banks, Universal metering etc.) there might be an opportunity for leapfrogging. In addition, I introduce a new idea (preliminary thoughts only!) on the directionality of information flow for power distribution management