Interesting times. Challenging times. Confusing times. The electricity industry and its regulators are now inextricably meshed in a tangle of interconnected reforms. With 50 states as laboratories, the process is accelerating. There is no going back. But which way is forward?
The old model of a closed system of vertically integrated electric utilities offering bundled service has been discarded in theory, and is being dismantled in practice. Consensus on a new model has not yet emerged. Parallel developments in FERC proceedings, state inquiries, and industry restructuring are lurching forward without recognizing fully how each part affects the others. There is a need for a greater sense of urgency here: Fundamental connections among several pieces of the overall puzzle must be recognized soon and incorporated in the reforms if the promise of open competition is to be realized, and the economic gains achieved.
The key lies in understanding the role of transmission and the requirements for efficient competition. The principal challenge today falls to the industry, and especially to the existing utilities. The FERC has broken through many barriers in its recent Notice of Inquiry (NOI) on alternative pooling institutions1 and its Notices of Proposed Rulemaking (NOPR) on stranded assets and transmission access.2 The various notices ask many questions and even suggest a few default answers. The FERC goes far (em perhaps as far as the FERC can go by itself (em but not far enough. The default proposals fall short of meeting the requirements for an open, efficient, competitive electricity market in a network.
Breakthrough or Breakdown?
Many utilities were so pleased with the FERC principles on stranded assets that they may have stopped reading the rest of the "Mega-NOPR" on transmission access. However, the industry would be wise to read the treatise with some care. There is a case to be made that the FERC process and the default proposals for transmission access have inadvertently headed down the wrong track. Having crafted an important breakthrough on the stranded asset principles, the FERC may have presented us with a work in progress on transmission access. If not stopped and redirected, this train may be headed toward a breakdown.
The key fault is not something that is there but something that is missing, something implicit that should be made explicit. The key implicit assumption underlying the proposal is that the nonprice terms and conditions for transmission access and service can be defined independently of the institutional structure and market pricing provisions. In other words, that transmission is like any other product or service. After all, we can define the characteristics of a bushel of wheat and its transport between locations without needing to define simultaneously the organization of the wheat market. We know a bushel of wheat when we see it, and just how many bushels are in a carload; so too with electricity and transmission (em at least that is the assumption.
This implicit assumption is found throughout the Mega-NOPR. Most striking is the companion inquiry into Real-time Information Networks (RIN),3 which imposes an accelerated schedule requiring comments to be filed 60 days before the corresponding deadline for the Mega-NOPR. [Editor's Note: Shortly before press, the FERC extended the RIN comment deadline 30 days to July 6, 1995.] In this RIN proceeding, the FERC expects to focus on "what information must be made available to transmission customers." The FERC wants the RIN rules put in place early, before the effective date of the open-access rule. But the FERC gives no indication that the information needed might depend on how transmission is defined, how prices are set, or how the electricity market is organized.4
The RIN project leads to a
second assumption (em that there
is an obvious way to define
transmission services. Without much discussion, the Mega-NOPR launches into references to "firm" and "nonfirm" transmission, "point-to-point" and "network" service, "wheeling through," "transmission capacity" and so on. These terms are not defined in the document in ways that relate to the technical requirements of transmission. They come from an underlying model of transmission service that is based on the "contract path" (em that electricity moves along a specific contract path from source to destination.5 The FERC recognizes the many deficiencies of the old contract-path model.6 Nevertheless, the Mega-NOPR effectively reverts to familiar terms and definitions that make sense only in the context of a contract path or a network without constraints. The tension is palpable
as the FERC condemns on one page what it implicitly accepts on another.
The internal tension appears again when the Mega-NOPR asks for comments on pools and dispatch services that complement or use transmission services. The apparent intent is to provide open access to these other essential services. A reader could easily conclude the FERC assumes that transmission services can be defined and offered separate from or in addition to other services such as economic dispatch.
If this underlying assumption is not correct, however, the entire strategy for approaching transmission access on the Mega-NOPR will be called into question. The farther we go in defining the terms and conditions for transmission under the implicit, old, rejected model of the contract path (em all the while creating presumptive property rights (em the more difficult the repairs when the train breaks down.
Contract Path Dead End
The definition of terms and conditions for transmission access can depend in important and, unfortunately, complicated ways on the organization of the institutions and the markets. The problems are many, but focusing on the definition of the transmission service of moving megawatts makes the point.
Under the old model of the vertically integrated utility, the "contract path" provided a workable definition of transmission service. The theory was that service would be provided as though the power actually flowed along a designated path in the network. When the power could not flow, the service was interrupted. Higher priority power could flow, and this would be firm service, except when it too was interrupted. And so on.
Most of us never knew or cared anything about this contract-path theory, and regulators could act as though the model was the reality. Of course, the utilities knew that this model had very little to do with what actually happened in the power grid. But as integrated monopolies it was easy for them to manage the few problems and handle the cost-shifting that was often implicit rather than explicit. In the new world of the competitive electricity market, however, this happy accommodation will not survive.
Under the contract-path model, and the assumptions of the Mega-NOPR, it would be possible to provide information in the RIN about the capacity of the various paths and the scheduled usage of the paths. The capacity would be defined in industry parlance as the "interface limit" for a set of lines, if not an individual line. Presumably, any user of the transmission system could look up the available capacity on an interface and commit to use some of that capacity independently of any consideration of the limits on other interfaces elsewhere in the system. The new power flow could be identified and the change on that interface recorded. Moving megawatts of electricity would be much like moving bushels of wheat.
How close is this stylized model to the real world? And how much would the differences matter? The answers are that the model is not at all close to the real world, and the differences matter a great deal. An interesting case is when the transmission system is congested. When the system is used only a little, anything can be done and the contract-path fiction can be accommodated. However, a constrained system leads to a dramatic result totally at odds with the contract-path model.
The system operators in the eastern interconnected grid regularly conduct joint studies of the transmission transfer capabilities of various interfaces. One of these exercises was conducted by the VEM Study Committee, which examined the impact of various power transfers under peak-load operating conditions.7 A central task was to evaluate the impacts of a power transfer across one interface on the transfer capabilities across other interfaces. For example, what would be the impact of a 1,000-megawatt (Mw) transfer from the Virginia-Carolinas (VACAR) region to Baltimore Gas & Electric and Potomac Electric Power?
The contract-path model assumes no impact on the transfer capability of other interfaces. As Figure 1 illustrates, however, the actual effects elsewhere would be far from zero, and certainly not negligible. The impacts would range from a gain of 50 Mw to a loss of 2,400 Mw, depending on the locations of the other interfaces. Clearly, parties quite distant from the transaction would experience major effects, sometimes larger than the originating transaction.
The complex network interaction or "loop flow" effect is caused by the nature of the highly interconnected grid and the current state of technology governing power flows. There are many interacting, nonlinear constraints that limit operations in power systems. The reduction to "interfaces" is a simplification used for network management in a highly coordinated system.
Further, the problems arise in any interconnected grid, not just in the highly networked system in the eastern part of the United States. Consider, for example, the simplified map of southern California that illustrates the location of major power plants, loads, and transmission lines (opposite page). The schematic includes three interfaces with associated maximum transfer limits: The East of River (EOR), with a maximum of 5,700 Mw; West of River (WOR), with a maximum of 8,206 Mw; and Southern California Import Transmission (SCIT), with a maximum of 16,974 Mw.
Under the contract-path model, presumably it would be possible to post these interface capacities and allow individual utilities or users to make decisions on how much capacity to use on each interface. In principle, the participants might assume that they could use both the 5,700 Mw on EOR and the 16,974 Mw on SCIT simultaneously. Unfortunately, the indicated capacities are not all achievable simultaneously. Actual use of the system imposes further limits that are summarized in the "nomogram" of Figure 2. This figure indicates the net effect of limits on simultaneous flows on the EOR and SCIT interfaces. Due to the interaction of load patterns with a number of physical limits, such as stability and voltage control, the allowable flow on one interface cannot be determined without knowing the flow on another. Furthermore, the limits on the flows depend on other factors such as the "inertia" of the available power plants operating in southern California and the status of the nuclear units at Palo Verde.
The interactions are complicated and large. To achieve the full SCIT limit, the EOR capability must be reduced from 5,700 to 700 Mw. Or, to use the full EOR limit, the SCIT flows must be cut in half. And when we note that the flows over the EOR would be counted again in the SCIT flows, the non-EOR imports across the SCIT could be reduced by as much as a factor of seven.
There are related dangers here, with ample worries for everyone. New entrants to the market should fear that the incumbents would decide to guarantee a path-based right under a wide range of circumstances by defining a very low transmission capacity, all of which is currently committed. Then new transmission capacity could be obtained only through expensive expansion, or not at all. For the incumbents the danger is that the larger nonsimultaneous limits may be allocated by regulators, with the cost of meeting them under different conditions imposed on the incumbents. At a minimum, incumbents would carry the burden of proof in demonstrating that capacity sometimes used would not always be available. For the regulator, the concern should be that capacity rights might be allocated in ways that artificially constrain available dispatch, increasing the cost due to congestion as the system operators stumble to keep up with the information that for every 1 Mw on the EOR interface someone may have to back off 2.13 Mw on the remaining SCIT flows.
as a Way Forward
Clearly it is impossible to identify separately the capacities of individual paths and then allow third parties to make their own decisions on how to use those paths. The real system doesn't work that way. The real system is a network that requires careful coordination and may behave in ways that have nothing to do with moving power from one location to another along a path.
In practice, utilities know
this well and take a network
perspective in actual operations. For example, rather than having San Diego Gas & Electric, the Los Angeles Department of Water and Power, Southern California Edison (SCE), and other users make independent decisions on power flows, they turn management of the SCIT nomogram over to SCE, and work through SCE in scheduling their power. This is a workable network-based system, but it is far from the contract-path based model.
If a competitive system is to be built upon specific performance and decentralized decisions (em where the contracted power actually flows based on the choices of the participants (em then the transmission services may need to be defined in terms of the parameters of the SCIT nomogram and the many other constraints that operate simultaneously. Or, if a pool-based network approach is embraced (em where the power flows according to the preferences of the participants, but through the choices of the dispatcher (em it would be possible to define transmission services in terms of financial contracts that convert the complicated interactions into locational price differences and simple financial settlements.
In each case, however, the unbundling of services and opening of access needs an explicit network model, not the implicit embrace of a contract-path fiction. The same problems extend to the unbundling of ancillary services.
Although never perfect, it is usually possible to separate the cost of reactive power support from the cost of spinning reserve, from the cost of frequency control, and so on. But for many services, no mechanism is available to identify the transaction requirement. The services are "joint" or "network" services that cannot at present be separated by individual transaction.
Leaping the Chasm
The approach in the Mega-NOPR is understandable. The FERC is best able to provide the framework and the incentives, but in the end it is up to the industry to propose and the FERC to dispose. Without a well-developed alternative, the institution reverts to what is familiar rather than what is real.
The challenge, therefore, is for the participants in the electric utility industry to come forward with the new approach based on a contract network that can replace the contract path. It is not possible to avert our eyes forever from the reality that the old model is dead, and the real problems of the network interactions cannot be wished away.
It is time to acknowledge a few basic facts. First, the parallel proceedings on alternative pooling institutions and transmission access are not truly separable; they are talking about the same thing, or at least two things that are fully intertwined. The separate proceedings should be brought together as one conversation. Second, transmission service is inherently, unavoidably, irrevocably, and importantly a network phenomenon. The definition, measurement, management, and pricing of transmission services must take a network perspective that integrates all these components. Third, this breakthrough is not beyond the ken of regulators or the industry. A workable consensus is near. The pieces have been implemented and tested elsewhere (em the most advanced version perhaps in Norway (em and proposed by a growing list of industry participants in the United States. The key to open, efficient transmission access in a network lies in coordination through a pool-based market that can support emerging competition. Network coordination provides the foundation for the voluntary pool-based model that could serve as the new consensus.8 Such a network-based approach would be more different than difficult, and would simplify the most vexing problems that won't go away. t
William W. Hogan is the Thornton Bradshaw Professor of Public Policy and Management at the John F. Kennedy School of Government, Harvard University, and director, Putnam, Hayes & Bartlett, Inc., Cambridge MA. This article encapsulates a presentation given at the 1995 Public Utility Research Center Annual Conference, and draws on work for the Harvard Electricity Policy Group and the Harvard-Japan Project on Energy and the Environment. Many individuals have provided helpful comments, especially Robert Arnold, John Ballance, Jeff Bastian, Ashley Brown, John Chandley, Doug Foy, Don Garber, Scott Harvey, Jere Jacobi, Paul Joskow, Jim Kritikson, Dale Landgren, Amory Lovins, Howard Pifer, Susan Pope, Larry Ruff, Michael Schnitzer, Irwin Stelzer, Jan Strack, Julie Voeck, and Carter Wall. The views presented here are not necessarily attributable to any of the above mentioned individuals, and any remaining errors are solely the responsiblity of the author.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.