2

Today’s Grid and the Evolving System of the Future

INTRODUCTION

This chapter describes the U.S. electric system as it now exists and discusses how it may evolve over the next several decades. First, the committee provides background on the physical, ownership, legal/regulatory structure, and operational characteristics of the nation’s electric system, with an emphasis on transmission and distribution infrastructure. The committee focuses on aspects of the national grid that are relevant for understanding electricity system resilience and the strategies employed to enhance it.¹ This overview of transmission and distribution also highlights the sensing, communications, and control systems that currently exist to support a variety of functions on the grid. Then, the committee describes the complex and dynamic forces driving changes in the electricity sector, both in the near term and the long term.² Finally, the committee discusses a variety of ways in which the system may change and some of the implications of these changes for the future resilience of the grid. Together, these conditions and trends set the stage for a subsequent discussion of threats to the system (in Chapter 3) and activities associated with each stage of resilience in the electric system (in Chapters 4 through 6).

Strategies to increase the resilience of today’s transmission and distribution systems need to accommodate possible future changes in its character, because most of the physical assets and other pieces of the infrastructure have long lifetimes. Planning to enhance resilience should take this into account, along with the often uncertain ways these systems might evolve over the coming decades.

Finding: Approaches to assure resilience should consider that components of electricity infrastructure have long lifetimes and that how the grid and its various institutions, technological features, legal structures, and economics will change is inherently uncertain.

ELECTRIC INDUSTRY STRUCTURE, ASSET OWNERSHIP, AND OPERATIONAL ROLES AND RESPONSIBILITIES

Since the 1930s in the United States, most electric service to households, businesses, and other customers has been provided by investor-owned or publicly owned electric utilities responsible for all elements of electric supply: generation, transmission at high voltage, and local distribution of power at low voltage. That said, in the first half of the past century the federal government promoted electrification and developed hydropower resources aggressively. This led to the federal government operating several electricity generation and transmission organizations, perhaps the most famous of which are the Tennessee Valley Authority in the southeastern United States and the Bonneville Power Administration in the Pacific Northwest. Figure 2.1 depicts the “bulk energy system,”³ comprised of central-station power plants and high-voltage transmission lines, and the local “distribution operations” that move power from the bulk system to end-use customers.

Several decades ago, most electric utilities were vertically integrated, meaning that the utility owned the power plants and/or contracts for power; owned or had rights to use high-voltage transmission lines that carry power from remote power plants to their local systems; and owned and operated the low-voltage distribution system to deliver power to consumers. State utility regulators (or, in the case of publicly

___________________

¹ Readers interested in a more detailed description might look at DOE (2017a), NASEM (2016), DOE (2015), MIT (2011), NRC (2012), and Bakke (2016).

² Readers interested in a more detailed description might look at MIT (2016).

³ The Federal Energy Regulatory Commission has approved the following definition of “bulk energy system” as developed by The North American Electric Reliability Corporation: “All transmission elements operated at 100 kV or higher and real power and reactive power resources connected at 100 kV or higher. This does not include facilities used in the local distribution of electrical energy” (NERC, 2016a). There are specific technical exclusions of certain facilities from this definition, but the 100-kV dividing line between bulk energy system (and transmission-level voltage) and lower voltage (and distribution-system-level voltage) is useful for our purposes here.

owned utilities, the governing boards of the local utility) set rates for vertically integrated utilities based on the cost of providing service. But nearly 20 years ago, a number of states and federal regulators began to move aggressively to break up vertically integrated utilities, separating the ownership of generation, high-voltage transmission, and distribution systems. In those states, only the distribution part of the system has continued to operate as a regulated monopoly.

As the electric system developed over the decades, investor-owned electric utilities in many parts of the United States merged so as to provide power to customers over larger and larger service territories. In other parts of the country, utilities serve smaller numbers of customers, particularly in rural regions where local electric cooperatives and municipally owned utilities continue to be the dominant providers of electric service. The result is today’s patchwork of local distribution utilities (Figure 2.2): thousands of electric utilities provide monopoly service within their local footprint but with a complex system of interconnected facilities that operates, in effect, as a single “machine” within each interconnection (NAE, 2003).

According to the Energy Information Administration (EIA), there are more than 2,000 utilities that own and/or operate some part of the generation, transmission, or distribution infrastructure in the United States (Table 2.1). More than 70 percent of end-use electricity customers are served by just 174 large investor-owned utilities, while the remaining customers are split roughly evenly between publicly owned utilities and electric cooperatives. Although these investor-owned and publicly owned systems are physically connected, their transmission and distribution systems often have different configurations, voltage ranges, and technology demands; are owned and/or operated by different parties; are subject to different types of regulatory oversight; and are frequently discussed separately.

These many utilities operate as part of three separate interconnected “synchronous” regions within the United States (and parts of Canada), as shown in Figure 2.3. Within

each interconnection, the utility systems are physically tied together by major transmission lines. The 60 Hz voltage and current waveforms are synchronized across the entire region, and power flows within each region according to the laws of physics. The three interconnections operate with only a few (asynchronous) direct current (DC) connections that allow transfer of energy between them. The major transmission lines serving the lower 48 states are shown in Figure 2.4. This figure also illustrates the strong synchronous connection with Canada for both the Eastern and Western interconnections, and the DC lines connecting the asynchronous Québec grid. The integrated North American power system mutually depends on close and continuing collaboration between the United States and Canada. And while there is also a connection to a small portion of Mexico within the Western Interconnection, that dependency is less significant for either country as most of the Mexican grid is a separate system.

TABLE 2.1 Breakdown of Utilities That Own and Operate Generation, Transmission, or Distribution Infrastructure

NOTE: Investor-owned utilities deliver 68 percent of electricity service to retail customers. Cooperatives, municipal utilities, and other publicly owned utilities deliver 13 percent, 12 percent, and 6 percent to retail customers, respectively. (As of 2015, 96 percent of electricity used by customers was sold through utility wires, with 4 percent generated on customers’ own premises.)
SOURCE: EIA (2016a).

Regulation of the electric grid takes place at two levels. The operations, cost allocation, and cost-recovery of the interstate transmission system, as well as wholesale sales of electricity,⁴ are largely regulated by the Federal Energy Regulatory Commission (FERC). FERC derives its authorities from the Federal Power Act (FPA), which was initially enacted in 1935 and has been amended multiple times. The second level of regulation occurs on distribution systems that deliver electricity to the end user. The terms and conditions of sales to retail electricity customers, including operations, cost allocation, and cost recovery for local transmission and distribution service, are subject to regulation by state regulatory agencies in those areas served by investor-owned

___________________

⁴ “Wholesale sales of electricity” are sales of power for resale to others, while “retail sales of electricity” are sales to ultimate, end-use customers. Retail sales are typically regulated by state utility regulatory agencies for investor-owned utilities (and by the governing entities of publicly owned or member-owned utilities).

utilities and by publicly accountable boards of public utilities.

This regulatory division between the federal government and the states over the higher- and lower-voltage portions of the electric transmission system first appeared in its current form in the early 20th century and has largely remained in place since then.⁵ Although seemingly straightforward, this division of authority is complex in practice and often gives rise to tensions. For example, although the FPA gives FERC authority over transmission service in interstate commerce and wholesale sales of electricity, the states have regulatory authority over siting of transmission lines (including the right to condemn right-of-way). Some states also retain regulatory authority over the costs of transmission as part of the bundled delivery of retail electricity (in vertically integrated states as described later). Further, many states have the ability to adopt a variety of tax, siting, environmental, and other regulatory policies that affect the mix of power plants in a state.

More than 20 years ago, the electric industry began to undergo pressures for structural change, in part owing to the experiences of deregulating other commercial sectors such as airlines, interstate trucking, and telecommunications. Additional impetus came from federal policies that supported the introduction of relatively small-scale, economical generating technologies owned by non-utility companies, which led to requirements that utilities open up their transmission systems for use by third parties (e.g., the Public Utilities Regulatory Policies Act [PURPA] of 1978). Efforts began in a number of states in the mid-1990s to separate the ownership of generation assets from ownership of the transmission system (the “wires”) and to create competitive wholesale electricity markets. A primary motivation in doing this was a belief that introducing market forces into the industry

___________________

⁵ As long-distance transmission lines emerged and utilities started to send power onto the grid across long distances, electricity began to cross state lines. Congress created FERC’s predecessor, the Federal Power Commission, in 1935 when it passed the Federal Power Act to address states’ inability to regulate interstate sales of electricity.

would result in lower costs to end users.⁶ In fact, creation of competitive wholesale markets in many regions of the country required that non-discriminatory access to transmission infrastructure be provided to all generators. After an initial flurry of “restructuring,” some states began to have second thoughts and decided not to break up their vertically integrated utilities.

Today, there is a patchwork of restructured and vertically integrated utilities across the United States. In much of the country, there are hundreds of non-utility entities involved in the power generation, system operations, power marketing, power trading, and other affiliated activities. The market participants in the electric regions serving two-thirds of the population in the United States are members of organized wholesale electricity markets where a regional transmission organization (RTO) (sometimes called independent system operators [ISOs]) operates the transmission system, prepares regional transmission plans for the market footprint, and conducts competitive product markets (covering energy, capacity, and/or ancillary services markets).⁷Figure 2.5 shows the boundaries of the current RTOs.

While retaining monopoly ownership of the distribution wires, several states also took steps to open up their electric systems to retail competition. In those shown in green in Figure 2.6, retail customers have the right to choose to buy electricity from competitive retail suppliers. Some states (shown in yellow) took initial steps toward allowing retail choice but then suspended it, while the remaining states (shown in white) did not introduce retail choice.

Across all of these areas, the specific terms and conditions of utility service, and any competitive supply, vary considerably. This makes it very difficult to generalize about industry structure across, and even within, states. At present this heterogeneous “electricity industry” reflects the varied choices that states and localities have made with regard to electric sector structure and regulation. The majority of states retain a vertically integrated structure, pursuant to which retail utilities maintain monopoly status with regard to the

___________________

⁶ In fact, in most cases, rates did not decrease (Lave et al., 2004; Blumsack et al., 2008).

⁷ As of 2015, these seven RTOs served 213.5 million, out of the total estimated U.S. population of 321 million (IRC, 2015; USCB, 2016).

generation, sale, and delivery of electricity. Many states that have vertically integrated utilities without retail choice (e.g., California and many states in the Northern Plains and Upper Midwest) nonetheless have utilities participating in RTOs.

As shown in Figure 2.6, one-third of the states decided to introduce retail choice, and a majority of the states’ utilities participate in the competitive generation markets administered by RTOs (shown in Figure 2.5), although the design of these markets varies across the seven RTOs.⁸ In some states without retail choice—for example, in Colorado—non-utility companies may own rooftop solar panels that are physically located on a customer’s building and sell that power to that customer. But, other such states without retail choice, such as Florida, do not allow anyone besides the utility to sell any form of electricity to consumers, although customers are able to install distributed generation on their premises. As a result of these variations across the states, the regulatory framework under which the electric grid operates takes on several forms. The FPA applies to the entire country but has differing impacts depending on which type of state-regional regulatory regime exists. This complicates the landscape in which the resilience of the interconnected grid is implemented.

The ownership of transmission infrastructure also varies widely across the United States. In some regions, vertically integrated utilities and large public power providers such as the Bonneville Power Administration and the Tennessee Valley Authority both own and operate the transmission infrastructure. In regions with competitive power markets, operation of the transmission system is delegated to RTOs/ISOs. These organizations may not own the transmission infrastructure under their control, but they are responsible for meeting reliability standards and conducting regional planning efforts, while assuring non-discriminatory access to transmission services for all generators and load-serving entities in the region.

With respect to reliability issues, FERC has responsibility for assuring adherence to mandatory reliability standards for the electric industry. FERC has delegated responsibility for developing reliability standards to the North American Electric Reliability Corporation (NERC), which had originally formed as a voluntary reliability organization following a large blackout in 1965 and is now the designated reliability organization in the United States. NERC develops industry-wide standards, submits them to FERC for approval, and

___________________

⁸ The only states that do not have any utilities participating in an RTO are Alabama, Alaska, Arizona, Colorado, Florida, Georgia, Hawaii, Idaho, Oregon, South Carolina, Utah, and Washington.

enforces approved standards in the industry. Thus, FERC does not develop reliability standards on its own. Compliance with NERC standards became mandatory with the passage of the 2005 Energy Policy Act (EPAct), and utilities and system operators now face substantial penalties for non-compliance.

Among many other things, NERC has defined the essential system functions necessary to ensure reliability in a framework that accommodates operational and structural differences across regions with and without competitive wholesale markets (NERC, 2010). Within each large region, there is a reliability coordinator with a wide-area perspective on system conditions necessary to ensure that the actions undertaken by one entity do not compromise reliability in another. Currently there are 12 reliability coordinators covering the Continental United States, much of Canada, and a small part of Mexico (Figure 2.7).

Under the purview of these reliability coordinators, more than 100 “balancing authorities” have responsibility for keeping generation and load equal at all times within smaller balancing areas. Regions with a history of tight coordination of operations and planning across utilities within the region, such as New England, New York, and the Mid-Atlantic region (e.g., Pennsylvania, New Jersey, and Maryland, the original location of the PJM territory), have only a single balancing authority, whereas the majority of reliability coordinators interact with multiple balancing authorities within their footprint. Box 2.1 has examples of transmission system oversight and operation that vary by region.

NERC directs several industry working groups and activities related to preparing for, riding through, and recovering from events with high impacts on the bulk power system. In addition, the Electricity Subsector Coordinating Council (ESCC), formed in response to recommendations from the National Infrastructure Advisory Council, provides a high-level forum for utility executives and federal decision makers to engage and maintain open communication channels in preparation for large-scale outages. To help reduce risks of cyber and physical attacks, for example, NERC operates the Electricity Information Sharing and Analysis Center (E-ISAC), which disseminates information and alerts to electric industry and government representatives, conducts training exercises, and also maintains the Cyber Risk Information Sharing Program that covers nearly 80 percent of operators of the bulk power system. Through the Spare Equipment Working Group, NERC maintains a database of

system components, particularly large transformers, which are available to participating utilities should their assets be physically damaged (NERC, 2011). Similar programs are maintained by industry trade organizations, such as the Edison Electric Institute’s (EEI) Spare Transformer Exchange Program and the Grid Assurance™ initiative recently launched by the private sector. Parfomak (2014) has prepared an excellent review of the issue of spare transformers for the Congressional Research Service. This report makes it clear that, while the past few years have seen progress, there is still much that needs to be done. The committee returns to the issue of replacement transformers in Chapter 6.

For many years, electric utilities have widely employed mutual-assistance agreements at both the transmission and distribution level to facilitate sharing of skilled workers and equipment to speed restoration efforts following outages. Typically restoration teams are composed with at least one local utility worker so that system-specific and regional knowledge is available on every team. After Superstorm Sandy, EEI developed a National Response Event Framework for pooling resources and coordinating restoration at the nation-scale from outages that overwhelm regional resources (discussed further in Chapter 6).

Thus, a hallmark of the U.S. electric system is that there are a myriad of bodies engaged in the ownership, planning, operation, and regulation of different elements of the system. Although the system itself operates as if it were a unified and coordinated machine, that occurs in spite of—or in the context of—a system in which the many component parts are subject to varied sets of institutional, legal, cultural, and financial incentives and penalties. Asset owners and operators must, and do tend to, operate with awareness of the fact that their systems can be impacted by events and developments occurring on other parts of the machine.

Finding: The “electric industry” is different across different parts of the United States in ways that reflect the varied choices that states and localities have made with regard to electric sector structure, asset ownership, and regulation. The specific terms and conditions of utility service, power system planning and operations, and transmission planning vary considerably, making it difficult to generalize about industry structure across and within the states. This complicates the landscape in which the issue of resilience of the interconnected grid must be addressed.

PHYSICAL STRUCTURE AND OPERATION OF THE HIGH-VOLTAGE TRANSMISSION SYSTEMS

Physical Structure

Most of the electricity supplied to today’s bulk power system is generated by large, central generating stations, often located far from population centers. Roughly one-third of the U.S. electricity supply comes from power plants that use natural gas, and another one-third comes from coal-fired generation. This reflects a significant increase in gas-fired generation in recent years, up from just 10 percent in 1990 (Tierney, 2016a). The fraction being generated by coal plants has fallen in large part because of competition from low-cost natural gas. Slightly less than 20 percent of generation comes from large nuclear plants. This share has been shrinking slowly, again because of competition from low-cost natural gas (and, to a lesser degree, flat demand and entry of renewable energy technologies) and the high cost of nuclear plant life extension. Hydropower produces 6 percent of the total U.S. power supply, with other renewables accounting for 7 percent of supply—most of that coming from wind (EIA, 2015). While power provided by large-scale wind and solar projects and from equipment such as solar panels located on customers’ premises is rapidly growing, it still constitutes a relatively small share of the total supply. These national averages do not reflect that some systems, such as those in California and Hawaii, have much higher percentages of distributed generation and intermittent renewables.

Hundreds of thousands of miles of transmission lines operate in interconnected networks across the United States, which carry alternating current (AC) electricity. Example voltages include 115, 230, 345, 500, and occasionally 765 kV. A few long-distance point-to-point lines use high-voltage direct current (DC) transmission.⁹ Electricity moves through the transmission system following the laws of physics and typically cannot be controlled precisely without expensive equipment.¹⁰ The bulk power system relies on large step-up transformers to convert electricity generated at central generating stations to high voltages; this allows for more efficient transmission of power across long distances because there are lower resistive losses of power at higher voltages.

Within the three U.S. bulk-power transmission interconnections, generators operate synchronously at 60 Hz. Large-scale electricity storage is relatively rare;¹¹ thus, power production and consumption must be kept in balance near instantaneously by increasing or decreasing electricity generation to match changing demand as customers increase and decrease their electricity use. In some areas, in addition to changing the amount of power being generated, grid operators use demand response (DR) programs and technologies to reduce certain loads in lieu of providing more generation. Maintaining the stability of this complex and dynamic

___________________

⁹ Direct current transmission is used selectively in the United States as a way to transfer power between asynchronous interconnects, occasionally to transfer bulk power over long distances (e.g., from the Pacific Northwest to California and from Labrador to the Northeast United States), and for underwater transmission (e.g., between Connecticut and Long Island and from offshore wind farms).

¹⁰ Technologies that allow control of AC power flows include phase-shifting transformers and other emerging power electronics-based flexible AC transmission system devices that are becoming more available and giving operators more control than ever.

¹¹ At present, the primary form of large-scale storage capability resides in hydroelectric pumped-storage facilities.

interconnected electric system is an immense operational and technical challenge. Nonetheless, this balancing act is successfully accomplished around-the-clock throughout the grid but not without the complex array of tools, techniques, systems, and equipment dedicated to the task.

The high-voltage transmission network enables power to travel long distances from generating units to substations closer to local end-use customers where the voltage is stepped back down and sent into the distribution system for delivery to consumers. Many of the approximately 15,000 substations have minimal physical protection, exposing them to natural hazards, vandalism, and physical attacks (NERC, 2014). Given that there is no standard design for substations, and especially for the transformers they contain, repairs and replacements of custom-designed facilities can be costly and take many months or even years to complete.

Most power outages occur on the local distribution system. Outages are less frequent on the transmission system. However, when outage events happen on the transmission system, they tend to result in wider impacts and can impose greater costs. Several of the largest outages—introduced in Box 1.1 and listed in greater detail in Appendix E—have resulted from operational or control-system errors followed by equipment tripping off-line due to close proximity with vegetation, as was the case with the 2003 blackout. Given the underlying network configuration of the high-voltage grid, system imbalances caused by events in one place can propagate across the transmission system near instantaneously, with the risk of causing cascading blackouts that impact customers hundreds of miles from the site of the initial disturbance.

Finding: Given the interconnected configuration of the high-voltage grid, events in one place can propagate across the transmission system in seconds or a few minutes, potentially causing cascading blackouts that can affect customers hundreds of miles from the initial disturbance. Thus, outage events on the transmission system can result in large-area impacts.

Sensing, Communication, and Control in the Transmission System

If electricity generation and consumption are not kept in balance, frequency will begin to rise or fall depending on whether there is a surplus or deficit of generated power, respectively. Deviations of voltage or frequency outside of relatively narrow boundaries can lead to physical damage to equipment and can increase the probability of a large-area cascading blackout. System operators depend upon various communications and other systems—for example, supervisory control and data acquisition (SCADA) systems in conjunction with software-based energy management systems (EMS)—to monitor the operating status (or state) of the transmission network and to control specific grid components to maintain stability. These systems rely on various sensors located primarily at substations (and, to a lesser extent, on transmission lines) to collect and transmit a wide variety of data, including voltage and current characteristics at specific geographic locations; environmental variables such as temperature, wind speed, and ice formation; and measures of asset health such as transformer oil temperature and dissolved gas levels (PNNL, 2015).

Autonomous local controls (called “governors”) at individual generators that boost power output proportional to declining system frequency (and vice versa) are fundamental components of system control responsible for regulating system frequency. The rotational inertia provided by spinning generators and some loads in each interconnection determines the rate of frequency change. On a slower time scale, the 60 Hz frequency is regulated by each balancing authority re-dispatching generation every few seconds through a wide area control scheme called automatic generation control.

Protective relays on the transmission network locate, isolate, and clear faults by triggering the appropriate circuit breakers to disconnect at-risk parts before the system becomes unstable and damage results. Depending upon their vintage, protective relays may be electromechanical (the oldest), solid state, or programmable and microprocessor based. They can act and take effect within tens or hundreds of milliseconds. To maintain acceptable voltage across long distance transmission lines, devices such as capacitor banks and static volt-amp reactive¹² compensators are used to control voltage.

A complex system of communications infrastructure is essential to the reliable operational performance of the electric grid, and this dependence is growing. There is, however, wide variation in the sophistication and speed of communication technologies used across the nation’s varied electricity systems, with equipment ranging from twisted wire, to wireless, to rented telephone line, to fiber-optic cable dedicated for utility use. The control of electricity systems is inherently challenging both because changes in the electricity system can occur very rapidly and because control needs to operate over time scales that range from milliseconds to multiple days.

To help system operators maintain system reliability, power systems have sensors, communications, and software that automatically perform analyses so as to constantly monitor the state of the electric system. The overall monitoring and control systems for transmission networks include displays and limit checking of all measurements for operators. A principal tool known as the State Estimator filters the various measurements and estimates the operational characteristics of the power system at regular intervals (e.g.,

___________________

¹² Delivered power is the product of voltage and current. In AC systems, only that portion of the current waveform that is in phase with the voltage waveform produces power. However, the out-of-phase current does flow in the lines and causes losses, so utilities strive to keep voltage and current waveforms in phase as close as possible.

every 30 seconds, although the time period used to be longer and continues to get shorter). This helps provide real-time assessments of system conditions that might not otherwise be observable by operators and improves their situational awareness. These assessments also enable other real-time analytic tools that can alert the operator to possible contingencies that could endanger the reliable operation of the grid.

Maintaining the security of these communication networks is critical to the operational integrity of the electricity system. Conversely, the integrity of these other systems (e.g., the internet and communications technologies) depends upon the operational integrity of the electricity system. Conventional approaches to cybersecurity such as firewalls, security software, and “air gaps” (i.e., no connection between systems) are used by utilities to protect their systems from intrusion. However, such measures are being recognized as inadequate, and the growing likelihood that breaches will happen motivates increased emphasis on cyber resilience, including intrusion detection and post-breech restoration. The importance of such activities is illustrated by the 2016 cyber attack on Ukraine’s electricity infrastructure. It took grid operators many months to even recognize that their systems had been compromised, at which point it was too late to prevent substantial outages from occurring.

To date, NERC has mandated nine cybersecurity standards as part of the overall mandatory standards it has established for the electric industry. These critical infrastructure protection (CIP) standards address the security of cyber assets essential to grid reliability.¹³ In addition to the cybersecurity standards from the Nuclear Regulatory Commission, these are the only mandatory cybersecurity standards for any of the critical infrastructure sectors across the United States (NERC, 2017).

Finding: System operators depend upon SCADA systems in conjunction with software-based EMS to monitor the operating status of the transmission network and to control specific grid components to assure safe and reliable operation. Control is inherently challenging because it must operate over time scales that range from milliseconds to multiple days. Maintaining the security of power system communication networks and control systems is critical to the operational integrity of the electric system.

Finding: CIP standards dictate minimum cybersecurity protections for the bulk power system, and the electricity sector is the only critical infrastructure sector with mandatory standards. However, these standards do not apply to local distribution systems.

PHYSICAL STRUCTURE AND OPERATION OF THE DISTRIBUTION SYSTEM

Physical Structure

The electric distribution system moves power from the bulk energy system to the meters of electricity customers. Typically, power is delivered to distribution substations from two or more transmission lines, where it is converted to a lower voltage and sent to customers over distribution feeders. Although distribution system outages tend to be more frequent than those occurring on transmission facilities, the impacts of such outages are smaller in scale and generally easier to repair.

Most local distribution systems in the United States are physically configured as “radial” systems, with their physical layout resembling the trunks and branches of a tree. Customers on radial systems are exposed to interruption when their feeder (i.e., their branch) experiences an outage. In metropolitan areas, these trunks and branches typically have switches that can be reconfigured to support restoration from an outage or regular maintenance. When a component fails in these systems, customers on unaffected sections of the feeder are switched manually or automatically to an adjacent, functioning circuit. However, this still exposes critical services such as hospitals or police stations to potential outages, so these facilities are often connected to a second feeder for redundancy. In high-density urban centers, distribution systems are often configured as “mesh networks,” with a system of interconnected circuits and low-voltage equipment able to provide high reliability service to commercial and high-density residential buildings. Such mesh networks—found in Manhattan, parts of Chicago and San Francisco, and other high-density urban areas—provide multiple pathways through which electric service may be provided to customers.

Most distribution systems’ wires are located aboveground. However, areas with high population density, including some suburban areas, frequently locate electricity and other infrastructure underground. This provides some physical protection and reduces risks posed by vegetation, but it can make identifying faults and implementing repairs more difficult and increase the risk of equipment damage in earthquake and flood-prone locations. In less densely populated areas, distribution feeders are usually located aboveground, with smaller distribution transformers located on local utility

___________________

¹³ NERC has nine mandatory CIP standards related to cyber issues. These cover such things as reporting of sabotage (CIP-001): identification and documentation of the critical cyber assets associated with critical assets that support reliable operation of the bulk power system (CIP-002); minimum security management controls to protect critical cyber assets (CIP-003); personnel risk assessment, training, and security awareness for personnel with access to critical cyber assets (CIP-004); identification and protection of the electronic security perimeters inside which all critical cyber assets reside, as well as all access points on the perimeter (CIP-005); a physical security program for the protection of critical cyber assets (CIP-006); methods, processes, and procedures for securing critical cyber assets and other cyber assets within the electronic security perimeters (CIP-007); identification, classification, response, and reporting of cybersecurity incidents related to critical cyber assets (CIP-008); and recovery plans for critical cyber assets, relying upon established business continuity and disaster recovery techniques and practices (CIP-009) (NERC, 2017).

poles that step down to lower voltage for delivery to customers’ premises.

There is no single organization responsible for establishing or enforcing mandatory reliability standards in distribution systems, although state utility regulators and boards of publicly or customer-owned utilities often assess performance using quantitative reliability metrics and set goals for the allowable frequency and duration of system and customer outages. Typically, utilities collect data on the length and frequency of outages that result from events on the local distribution systems, and some utilities (particularly investor-owned utilities with encouragement from regulators) disclose this information to the public. However, there is wide variation across the states and the utilities within them with regard to their tracking, publication, and/or enforcement of local reliability indicators. In light of their role in approving rates and in deciding what costs and other investments can be recovered through rates, public utility commissions (and boards of publicly or customer-owned distribution utilities) have significant influence on the reliability, cost, and resilience of distribution systems, as FERC does at the bulk energy system level.

In recent years in some parts of the United States, distribution systems have also experienced substantial additions of distributed energy resources (DERs). DERs are electrical resources that are attached to the local distribution system, often behind a customer’s meter. Examples include rooftop solar panels, customer-owned batteries, fuel cell technologies, wind turbines, backup generators, and combined heat and power (CHP) systems.¹⁴ Although DERs account for a relatively small fraction of total generation nationally, their installation varies significantly from one state to another, with some local distribution systems (e.g., in Hawaii, California, New Jersey, and Arizona) seeing hundreds of MW of growth in installed capacity in recent years (DOE, 2017a). Because many DERs provide surplus power beyond the amount of electricity consumed on the customer’s premises, they inject power into a distribution system designed to operate in a one-way flow of power from the substation to the customer. (See “Near-Term Drivers of Change and Associated Challenges and Opportunities for Resilience” for a longer discussion of DERs and their implications for grid planning, operation, and resilience.)

Even with increasing numbers of consumers installing generating equipment on their own premises, and using the distribution system to access the bulk energy system when on-site generation is not available, it is unlikely that the majority will go entirely “off grid” in the near future. Although many technologies and service offerings are enabling an increasing number of customers to meet larger portions of their electricity needs with on-site generation, for economic, technical, and regulatory reasons most observers (and the committee) do not anticipate that the dominant customer profile will be self-sufficient and disconnected from the grid during the time frame of interest in this study (i.e., in the next two decades). Moreover, individual self-sufficiency is unfeasible for the majority of the population, and local distribution system planners have to plan to meet the uncertain loads of customers for the foreseeable future.

Finding: There is no single organization responsible for mandatory reliability standards in electric distribution systems in the United States. State utility regulators often set standards for the allowable frequency and duration of system and customer outages. In many cases, outages caused by major events are excluded when computing reliability metrics.

Sensing, Communication, and Control in the Distribution System

The technological sophistication, penetration of sensors, deployment of advanced protection devices, communications technologies, computing, and level of automation deployed by distribution utilities vary significantly across the United States. As in the case of transmission systems, distribution networks have been undergoing a transition from analog devices to digital. However, in many distribution systems, it is more difficult to justify large investments in modernization and digital controls, in part owing to factors such as customer density on circuits, circuit configurations, existing performance, and component age. Thus, many distribution systems still operate as they did when built after World War II. However, given the substantial investments (exceeding $25 billion annually [EEI, 2017]) under way in replacing aging distribution infrastructure, there is an opportunity to enhance the reliability and resilience of the distribution systems through incorporation of advanced technologies, and some distribution utilities have made extensive upgrades.

Protective relays located at distribution substations are used to sense faults, such as a downed wire, and in turn signal the feeder circuit breaker to open. Some feeders have switches that can detect and isolate faults, albeit less frequently (as discussed previously). Distribution laterals that extend from the main feeders have fuses installed that protect the main feeder from faults that occur on the lateral branch. Together, protection devices are critically important for maintaining public safety and for limiting the extent of an outage, in some cases preventing disturbances from cascading higher up in the system.

Each of these devices, relays, switches, and fuses are designed to operate in a coordinated manner. These distribution protection schemes are undergoing a similar analog to the digital transformation occurring on transmission systems. Over the past 20 years, electromechanical relays

___________________

¹⁴ Certain energy efficiency measures can function as DERs so long as they are dispatchable, meaning they can be turned on or off when needed by the utility. Other definitions do not emphasize that DERs be dispatchable—for example, FERC’s definition at https://www.ferc.gov/whats-new/comm-meet/2016/111716/E-1.pdf.

have increasingly been replaced with digital, and now communicating, software-based relays as old equipment reaches end-of-life or when new substations are constructed. Similarly, switches on some feeders have been replaced with more advanced and automated switches when it is cost-effective and justifiable. Protective fuses also have digital communicating alternatives, but these are still largely in demonstration studies to evaluate cost-effectiveness and applicability.

Beginning in the 1990s, many utilities selectively installed SCADA on distribution systems for feeder breakers, midpoint reclosers, and back-tie switches (as well as capacitor bank controls), along with distribution management systems to operate these devices. These first-generation automation systems allowed utilities to operate circuit breakers, switches, and components remotely, which previously required personnel in the field. By sectionalizing circuits in half, these early systems allowed more rapid restoration of the faulted half of the circuit. Such systems have been implemented by many utilities in metropolitan areas where high customer densities enable cost-effective applications.

More recently, a second generation of distribution automation technologies has been adopted. Outage management systems (OMS) that provide greater visibility into distribution circuits and support operators in making restoration decisions have been deployed over the past decade. Some utilities have implemented advanced automation technologies that locate faults, isolate faulted sections, and automatically restore remaining sections to service. Similar to first-generation automation systems, these systems are typically cost-effective only in areas with high customer density per mile of line and on overhead lines with exposure to environmental conditions that reduce reliability and impair restoration.

Although at present these technologies have only been implemented on a fraction of distribution systems across the country, continued deployment of distribution substation SCADA and first- or second- generation automation has the potential to improve the reliability and resilience of the nation’s distribution systems, albeit if implemented selectively and as part of a long-term improvement plan. For example, select utilities in areas with significant exposure to environmental threats (e.g., Southern Company in the southeastern United States), or with the need to have greater visibility and control over DERs (e.g., Southern California Edison), have installed or are pursuing advanced automation technologies for automatic reconfiguration of feeders based on outage and load/local generation conditions. However, it is unlikely that these second-generation automation technologies will be deployed in lower-density rural areas or in newer underground systems, as the potential benefits do not typically justify the increased costs.

Compared to transmission systems, which have greater deployment of sensors and therefore provide operators with much better awareness of system behavior and operation, often local distribution utilities only monitor circuit breaker status and measure feeder current and voltage as they leave the substation, and not at other locations on the circuit. However, some utilities installed automation sensing and fault current indicators on feeders themselves, although this level of monitoring is uncommon. Thus, most distribution utilities continue to rely on customer calls to assist in the location of faults. In the most rudimentary cases, utilities without distribution substation SCADA use customer calls to report outages and direct service restoration and repairs.

Utilities have yielded significant benefits from first-generation distribution automation, where cost-effective, but second-generation automation systems are still early in adoption (DOE, 2017b). One utility that adopted second-generation automation with the help of federal demonstration grants reported significant reductions in the severity and duration of outages, as well as economic and operational benefits (Glass, 2016). Of course, actions that increase automation, reliance on software, and communications infrastructure also add complexity and can inadvertently increase a utility’s exposure and vulnerability to cyber attack.

Within the past decade, utilities have completed more than 60 million advanced metering infrastructure (AMI, sometimes also called “smart meter”) installations across the United States. These investments were greatly accelerated by incentives arising from funding available in the 2008 American Reinvestment and Recovery Act. Figure 2.8 shows the percentage of electric meters with AMI by state. In distribution systems where it has been installed, AMI can provide information to assist in identifying the extent and location of customer outages, as well as the primary benefit of reducing the cost of meter reading. However, the outage data from AMI systems tend to be of poor quality and inconsistent for use in real-time fault identification and initial restoration. This is in part because the messages sent to operators are a “last gasp” from a meter losing power, and often the message itself cannot get back to the operations center as the communications network also loses power (most AMI systems installed are based on radio frequency mesh communications networks). As a result, most AMI systems today are used to validate that electricity service to customers has been restored and for postmortem analyses. More advanced AMI systems, which are available today, have addressed this issue and will be able to support real-time operational restoration and improved communication with customers. Furthermore, to take full advantage of AMI, utilities must make substantial investments in database management and analysis software to utilize the large amount of data flowing back to operators.

Deployment of advanced meters has been met with mixed reactions. Some state regulators remain skeptical of the benefits of AMI or contend that equivalent benefits can be achieved at a lower cost to customers (Reuters, 2010; AEE, 2015; NJBPU, 2017). Some customers have been suspicious of technologies that they view not only as expensive, but also

as potentially dangerous for their health¹⁵ and for the security of their private data (Karlin, 2012; Spence et al., 2015). AMI roll-outs in some communities have experienced backlash for these reasons, although other AMI deployments have been much smoother.

Inverters convert the DC signal produced by solar panels or batteries to the AC power used on the distribution system and serve as the interface between many DERs and the distribution system. While the main task of an inverter is as an electric power conversion device, modern technology permits inverters to perform a broader array of ancillary tasks, which can be leveraged in power conditioning to support the grid in various ways (these are sometimes referred to as “Smart Inverters”). Currently, inverters operate with a spectrum of capabilities—for example, some are able to stay connected and ride through disturbances (and in some cases can contribute to solutions), while others automatically disconnect during a disturbance. Interim standards issued by the Institute of Electrical and Electronics Engineers (IEEE) allow for such “ride through” of disturbances, and FERC now requires this capability. These standards remain under revision (IEEE, 2013).

Currently, relatively few of the inverters installed on the system can provide the local utility with visibility into the power injection of the DER into the grid or the ability to control it when necessary. At some point in the near future, when technical standards catch up with technology, it is possible that inverters will have the capability to communicate with utilities and system operators. This can be further leveraged to enhance system resilience under abnormal situations—for example, by changing inverter settings on the fly for adapting to changing grid conditions. Additional details are provided in the discussions in Chapters 4 and 5.

Finding: There is wide variation across the United States in the level of technological sophistication, penetration of sensors, deployment of advanced communications technologies, and level of automation deployed by distribution utilities. Many utilities, particularly in metro areas with overhead infrastructure, have invested significantly in first-generation automation over the past 30 years. Where cost-effective,

___________________

¹⁵ While the field strengths are miniscule, the concern is with the possibility of health consequences from exposure to the RF communication associated with the AMI. Similar concerns are expressed by some people about a wide range of RF sources in the world today.

more advanced automation is beginning to be implemented to enhance reliability, resilience, and integration of DERs.

Finding: Actions that increase automation and reliance on software and communications infrastructure also add complexity and can inadvertently increase a utility’s exposure and vulnerability to cyber attack. This is particularly acute with regard to DER integration.

Keogh and Cody (2013), researchers with the National Association of Regulatory Utility Commissioners (NARUC), explain the following:

Thus, because the existing reliability metrics used to inform regulatory decision making are inadequate for informing resilience investments, continued research is needed to develop analogous metrics for electricity system resilience. Some regulators have begun to consider how resilience objectives should be incorporated by utilities in their jurisdictions, with several prominent examples promising to transform the electric industry today. In response to Superstorm Sandy, for example, New Jersey regulators approved more than $1 billion in storm-hardening investments for critical substations and building additional distribution circuits for greater redundancy (NJBPU, 2015).

Finding: The decisions made by state public utility commissions and the boards of public or customer-owned utilities have significant influence on the reliability, cost, and resilience of distribution systems. The committee agrees with a NARUC analysis that concludes that techniques for guiding and approving reliability investments are inadequate for resilience.

METRICS FOR RELIABILITY AND RESILIENCE

Reliability Metrics Are Relatively Mature and in Widespread Use

Reliability has long been a component of utility planning and operation, and there are many mature metrics to quantify reliability and evaluate potential reliability improvements associated with different grid investments. Reliability metrics are grouped into those applied to generation and transmission systems (e.g., adequacy, loss of load probability) and those for the distribution system, with common examples defined in Box 2.2. Metrics for generation and transmission are used by FERC and NERC, whereas oversight of reliability at the distribution level is left to state regulatory agencies. As previously discussed, ownership and operation of the U.S. electric system is characterized by a mixture of public, private, and cooperative institutions with different incentives and organizational structures, and these different institutions are regulated differently. Thus, different organizations are responsible for maintaining different packages of standards in different locations, some of which can only be attained through collaboration with others.

While reliability metrics are more established and widely used than resilience metrics, there remain many opportunities to improve their formulation and utilization. Although valuable, distribution system metrics that present average values lack details regarding the types of customers experiencing an outage and the severity of individual outage events. Thus, there is a need to increase the granularity of reliability metrics, and the Department of Energy (DOE)-sponsored Grid Modernization Laboratory Consortium (GMLC) is in the process of developing metrics for distribution reliability with greater spatial and temporal resolution (GMLC, 2017). Another critical opportunity for improvement is to better connect reliability metrics to the economic benefits of more reliable service, which requires an understanding of how different customers value reliable electric service.

As society becomes ever more dependent on continuous electricity supply, and the technologies and institutional structures employed to provide that service evolve, it is important to rethink the system’s reliability criteria. To the extent that electricity supplies become more distributed, micro-sized local supply communities may take care of their own unique local needs; but to the extent that a significant component of supply is provided over a regional power grid, all users share equally in that bulk supplier’s reliability (what is defined as a “public” good by economists) and so some centralized authority is needed to set and enforce the reliability standard for that supply entity. That standard could be based and routinely updated on some systematic estimate of the value of its reliability (and resilience, too).

___________________

¹⁶ The authors also explain, “If an investment avoids or minimizes service interruptions in the absence of an extraordinary event, it is just an everyday reliability investment, and the means already exist for utilities and regulators to thoroughly consider it. An important point . . . is that resilient infrastructure does more than one thing well, because a resilience investment needs to pay for itself and create value for ratepayers, even when it is not being used” (Keogh and Cody, 2013, p. 5).

It is important to note that reliability metrics provide only limited insight about resilience. A survey of publicly owned utilities in 2013 indicated that two-thirds of the responding utilities excluded outages caused by major events when calculating their performance on reliability metrics (APPA, 2014).¹⁷ Thus, planning, operational strategies, and technologies used to reduce impacts and expedite recovery from large-area, long-duration outages may have no impact on a utility’s performance measured by reliability criteria.

Development of Metrics for Resilience Lags Behind Those for Reliability

Unlike reliability, there are no generally agreed upon resilience metrics that are used widely today. This is in part because there is not a long history of large-area, long-duration outages that can be analyzed to guide future investments (which is the case for reliability). Nonetheless, the electricity sector is arguably more advanced in considering and evaluating resilience than other critical infrastructure sectors. There are myriad resilience metrics proposed in research and most remain immature (Willis and Loa, 2015). Some recent analyses have proposed resilience metrics based on concepts like resistance, brittleness, and dependency. Following the resilience processes introduced in Chapter 1, Kwasinski (2016) proposes that resilience is an attribute with four distinct metrics: (1) withstanding capability, (2) recovery speed, (3) preparation/planning capacity, and (4) adaptation capability. A study at Sandia National Laboratories lays out a broad framework for developing resilience metrics, frequently in combinations, and for valuing their respective contributions to overall customer value (SNL, 2014). Furthermore, individual utilities frequently establish their own metrics to guide decision making. For example, the committee was briefed by the Chicago utility Commonwealth Edison on metrics used in selecting optimal locations to site community microgrids,¹⁸ based on a weighted sum of measures of

___________________

¹⁷ Also, of the 180 utilities responding to the American Public Power Association survey, 87 percent collected outage data at the system level, 47 percent also collected data at the feeder or circuit level, and 31 percent collected data at the substation level (APPA, 2014).

¹⁸ A microgrid is an energy system consisting of distributed generation, demand management, and other DERs that can connect and disconnect from the bulk power system based on operating conditions.

customer criticality, historical reliability, projected capacity constraints, and measures of substation health.

As part of the GMLC metrics analysis, researchers from multiple national labs proposed a set of resilience metrics, shown in Table 2.2, that build on a resilience analysis process developed as part of the DOE Quadrennial Energy Review. Because many causes of large-area, long-duration outages have a low probability and their impacts are highly uncertain (e.g., based on the types of customers impacted, the exact tract a hurricane follows), the GMLC metrics analysis emphasizes inclusion of statistical measures of uncertainty alongside reporting of resilience metrics and all consequences are estimated as probability distributions.

Development of resilience metrics and methods to defining resilience goals, as well as comparison of alternative strategies for increasing resilience, remains an active area of research, and the committee believes more research and demonstration is required before the electricity sector can reach consensus on a set of appropriate metrics. Metrics often drive decision making. Establishing and building consensus around metrics is an important prerequisite for comparing resilience enhancement strategies and for evaluating their costs and benefits. Many of the technologies and strategies for increasing the resilience of the electricity system described in the following chapters are expensive, particularly when implemented on a large scale. Without consistent resilience metrics, large amounts of money could be spent with little understanding of actual resilience benefits and with much of this cost passed on to ratepayers.

TABLE 2.2 Example Resilience Metrics Proposed by the Department of Energy-supported Grid Modernization Laboratory Consortium

SOURCE: GMLC (2017).

Economic Valuation of Resilience

Metrics for resilience should not be selected merely because they can be quantified easily. In deciding what level of resilience is appropriate, it is important at a minimum to estimate how much a lack of electricity system resilience costs individuals and society. Thus in developing resilience metrics, it is essential to be able to link those measures to the value retained or added to society. Furthermore, market responses and/or survey results may provide inadequate measures of resilience since they have attributes of both a private and a public good (many neighbors share the same benefit). Likewise the services provided by most public or private regulated utilities are combinations of pure public and private goods. This is why standards and regulations are important to maintain and restore quality in electricity markets, which are not classical competitive markets with fully rational decision makers (Hirschman, 1970).

Thirty years ago, with most electric supply utilities vertically integrated, the customers knew who to blame for outages. If the overseeing public utility commission (PUC) did not set and enforce adequate reliability standards, the resulting public outcry often resulted in a government response including public pillorying and/or financial penalties assessed against the responsible utility. In some instances of major outages, the outcry extended to elected officials in state or federal government. The principal example is the 2003 blackout that led to EPAct of 2005, granting new authority to FERC to set reliability standards for the bulk power system and to assess penalties for non-compliance.

Developing and enforcing resilience and reliability metrics will become increasingly complicated as technologies and customer preferences evolve alongside changes in public policies regarding equity and environmental goals. The emergence of competitive markets in some areas of the country has altered the institutional structure of the industry, the nature and form of its regulation, and the structure of its financing. So while competition has replaced regulation in some segments of the industry as the means of ensuring reasonable price levels, maintaining the reliability of the whole system has become more complicated with divided responsibility. At the bulk power supply level today, reliability standards are still maintained, but this is often done through market mechanisms that induce sufficient prices for

adequate generation to be built at needed locations, as well as for generation operators to provide operating reserves and to be available to offer those services (provide adequacy), all as overseen by FERC. At the distribution level, state regulation (and public outcry) is primarily relied upon to sustain the reliability to end-use customers.

In the end, reliability and resilience are for the benefit of the customer and society, and all actions, including rules and regulations, need to reflect customer values. Although a consistent principle should be developed for the nation, cost-effective instruments are likely to vary widely. The application of the principle should take into account variations in climate, nature of hazards, socio-economic and demographic patterns, and the nature of customers (industrial, commercial, residential, essential public services, etc.), all of which may lead to different distribution-system configurations (e.g., there are mesh network designs in some densely populated areas, whereas less populated areas have radial distribution system designs).

No rule is effectively implemented without rewards or penalties assigned for adherence. For private goods, if there is truth in labeling and no hidden defects are possible, the market can take care of those incentives. In the case of public goods furnished by a unique provider in each location, assessing penalties for non-compliance can have pernicious repercussions if the service must be sustained. If compliance requires substantial capital investments, arranging financing can be challenging if the entity is under attack by its regulators and its next period’s earnings promise to fall because of the fines. If fines are pooled over a wide area of providers in order to support resilience and reliability investments, there is little incentive for the individual utility to provide reliable service. The nature of such problems will change if numerous local microgrids and community-based distribution consortiums become widespread. Furthermore, the shifting of reliability and resilience decisions to the local level also presents serious challenges for financing. One model might be parallel to the U.S. Department of Agriculture Rural Utility Service’s (RUS’s) funding of rural cooperative electricity suppliers.¹⁹ In the end, regardless of the form of the institution, reliability and resilience begins at home—at the distribution level with the customer.

Because electricity customers value both the reliability and resilience of the system, developing metrics and incentives (or disincentives) for utilities based upon resilience and reliability separately is likely to be sub-optimal. It is important that the possibility of trade-offs between resilience and reliability is integrated into metrics, and that the costs of supplying the sum of the measures do not exceed their combined value to customers and to society as a whole (SNL, 2014). At present, such an overarching valuation of the burgeoning number of reliability and resilience metrics does not exist to aid in the development of reasonable and enforceable standards.

In addition to developing better resilience metrics and using them to monitor and realize better outcomes, knowing much more about what individuals and society are willing and able to pay to avoid the consequences of large-area, long-duration grid failures is an important input to deciding whether and how to upgrade systems to reduce impacts of an outage. Much of what we know is anecdotal from looking backwards at such failures, such as from Katrina, Sandy, or the Northeast blackout of 2003. Most prior quantitative studies have only examined outages of much shorter duration. Willingness and ability to pay may differ substantially based on geography, electric customer class, and socioeconomic status. So work should proceed in parallel to develop better metrics and a better understanding of consumers’ and society’s willingness to pay.

Finding: While reliability metrics are relatively well established and widely used in electricity system planning and operation, the development of agreed-upon metrics for resilience lags significantly behind. Further, since there is currently no common basis for assessing the relative cost-effectiveness of the existing reliability metrics that differ by purpose, integrating the ongoing work on developing resilience metrics may lead to confusion and duplication in their implementation. Thus it may be difficult to evaluate, compare, and justify investments made to improve resilience and to assess progress made in enhancing both the resilience and the overall reliability of the grid.

Recommendation 2.1: The Department of Energy should undertake studies designed to assess the value to customers—as a function of key circumstances (e.g., duration, climatic conditions, societal function) and for different customer classes—of assuring the continuation of full and partial (e.g., low amperage and/or periodic rotating) service during large-area, long-duration blackouts.

Recommendation 2.2: The Department of Energy should engage the North American Electric Reliability Corporation, the National Association of Regulatory Utility Commissioners, the National Rural Electric Cooperative Association, and the American Public Power Association in a coordinated assessment of the numerous resilience metrics being proposed for transmission and distribution systems and seek to operationalize these metrics within the utility setting. That assessment should focus on how system design, operation, management, organizational actions, and technological advances are affected by those metrics. All metrics should be established so that their cost-effectiveness in bringing added value to the nation can be assessed. Complementarities between metrics should be identified, and double counting of their effects should be avoided.

___________________

¹⁹ The RUS provides loans and loan guarantees to help finance construction and operation of electric distribution and transmission systems (among other things) in rural areas. Electric cooperatives (and other utilities) may receive such financial support from the RUS (USDA, 2016).

NEAR-TERM DRIVERS OF CHANGE AND ASSOCIATED CHALLENGES AND OPPORTUNITIES FOR RESILIENCE

As described previously, significant transitions are currently under way in the power system and its associated institutions. Some changes result from market fundamentals including changing customer preferences, others from an array of state and federal policies, and yet others from technological innovations that offer both opportunities and new challenges for the grid, especially in terms of resilience. The future electric system will have a more complex array of central-station power plants on the bulk power system, as well as DERs behind customers’ meters or otherwise attached to the local distribution system. Many more players will use technologies and applications that can expose the grid to greater risk of cyber attack. These changes may both facilitate and complicate the development of greater reliability and resilience. Starting with a description of these various trends that are affecting the grid, this section discusses some of the implications of those trends for the resilience challenges its owners, operators, and users will increasingly face in the years ahead.

Power Market Fundamentals

The nation’s “shale gas revolution” began a decade ago and has contributed to a changing generation mix in many parts of the United States, particularly where coal-fired or nuclear generation have been major players. In combination with a decade of flat electricity demand (EIA, 2016b), loss of cost advantages for coal (Tierney, 2016a), declining costs for small-scale and utility-scale wind and solar generating technologies (Lazard, 2015), and controls on emissions of mercury and other toxic air pollutants, this has contributed to retirements of 49.3 gigawatts (GW) of coal-generating capacity since the year 2000 (EIA, 2016c). Most of these plants were older, relatively inefficient, and without modern pollution controls. Because of competition from low-cost natural gas and the high costs of plant life extensions, several nuclear plants have been retired in recent years with others facing premature closure (BNEF, 2016).

The vast majority (91 percent) of the 403 GW of generating capacity added since 2000 has been at gas-fired generating units (281 GW), as well as wind and solar installations (together, 87 GW) (EIA, 2016d). In 2016 alone, utility-scale wind, solar, and gas-fired capacity amounted to 93 percent of total generating capacity additions (EIA, 2016d). Another 2 GW of distributed solar capacity was added in 2015, which is the most recent year reported by EIA (EIA, 2016e). The changing electric generating mix is introducing new challenges for grid operators, who must keep generation and consumption balanced with a decreasing amount of baseload coal and nuclear assets and an increasing share of intermittent, non-dispatchable generating resources.

DERs differ from the large central generators that traditionally form the backbone of the grid in that DERs are much smaller, located closer to consumers, and often controlled in a decentralized fashion by local users themselves. The shift to DERs comes as a result of changes in technology, customer preference, and policy. Technologically, numerous new power supply, response, and control systems are emerging. At the same time, federal and state regulators, as well as others, are pushing for the adoption of DERs with a variety of goals that are described further in Box 2.3 and in the following section. As with almost any change in technology, these driving forces interact in many complex ways. Some of the changes in technology are purely exogenous, but most are responding at least partly to policy signals. These forces also interact with consumer preferences, as is typically observed with changes in other technologies. New technologies for local supply and power conditioning have seen early adoption by users who have a particularly strong preference for reliable power, such as hospitals and server farms.

Federal and State Policy Drivers

The federal government and most states have been active in adopting policies aimed at promoting the introduction of efficient and renewable energy technologies, controlling emissions associated with power generation, and fostering innovation and grid modernization. These policies, many of which are mentioned in Box 2.3, have impacted both the bulk power and local distribution systems. Importantly, but with notable exceptions, federal and state policies that have encouraged development of advanced technologies and DERs have been motivated by considerations of economic development, environmental impacts, or clean-energy goals, rather than by concerns for resilience and reliability.

While many of these federal and state policies have been directed toward regulated utilities, many have encouraged non-utility entrants to make investments, operate programs, and bring new technologies to the marketplace. Today, many of the devices (e.g., central-station power plants, rooftop solar installations and their accompanying smart inverters) attached to the grid are owned by third parties. There are many more actors affecting the operations of the grid, and grid operators and others need to take into account a wide variety of facilities and resources as they assure the operational reliability and security of the grid.

To gain a better appreciation of the state of DER and microgrid adoption in jurisdictions across the country, the committee sent a questionnaire to public utility commissions in all 50 states and the District of Columbia and received nearly 25 responses. The questionnaire sought anecdotal information about variations in deployment of smart meters, distribution automation, organized DR programs, CHP facilities, and questions regarding legal constraints on microgrids across the country. Answers called attention to wide differences in adoption of these technologies and views on their

potential to increase system reliability and resilience across the United States, as summarized in Box 2.4. Although not quantitative and not used to make any comparative statements, the answers received by the committee broadly align with previous studies done by FERC (2016b) and stakeholder groups (Gridwise Alliance, 2016).

Changing Time Scales for Grid Operators

Along with the changes to the fundamentals of the generation mix, the electricity power system is undergoing changes to the time scales for operations, especially in the area of power markets for restructured utilities. The future will see continued shortening of time scales for grid operations: data on system conditions come in on time scales under a second, and the dispatch of resources and market settlements happens every 5 minutes. The requirements for such rapid dispatch and analysis have impacted the tools used to manage the system, causing the energy management systems within RTOs to be custom built. The operational concerns of the collapsing time frames and the human interface are real. Though the resilience impacts of these changes are complex, these challenges motivated the committee to recommend research on improvements to system operator control rooms and the application of artificial intelligence to power system monitoring and control within Chapter 4. These concerns also help motivate overarching recommendations to improve the security and resilience of the cyber monitoring and controls systems within Chapter 7.

Industry-Structure and Business-Model Transitions

There are new industry structure and business model issues that are also in transition, with uncertainty about which direction they will take in the future (NASEM, 2016; MIT, 2016). Competitive forces, often stimulated by actions of federal and state legislatures and regulators, have prompted an array of new actors (e.g., non-utility generating companies and independent non-utility transmission companies), new institutions (e.g., RTOs and ISOs), and new issues subject to FERC regulation in wholesale electricity markets and the bulk power system. Most of these institutional changes have

already occurred. Unlike the bulk power system that has undergone significant restructuring and regulatory reform over the past decade, the structure and regulation of electric distribution systems has, until recently, experienced much less change. Thus, the committee considers that the largest changes to the structure of the electricity system in the future will occur within the distribution side of the system.

At the distribution-system and retail electric level, the relatively rapid emergence of DERs has accelerated pressure on regulators, utilities, and other stakeholders to address aspects of the traditional utility business model, which has supported grid investments largely through rates that recover significant quantities of utilities’ fixed costs through usage-based charges. All else equal, as new small-scale technologies generate power from customers’ premises and inject it into the grid (Figure 2.9), causing revenues from volumetric rates charged to customers to drop, utilities and others have begun to look for regulatory frameworks and new rate designs that assure that all customers pay their fair share of the costs of maintaining a reliable and resilient grid. The approaches under discussion across the country for the future roles of the local distribution utility include the “enhanced status quo,” the “network service provider,” the “market enabler,” and the “solutions integrator” (De Martini and Kristov, 2015; State

of New York, 2014; Tierney, 2016b; TCR, 2016). These new business models are relevant for resilience considerations in light of the fact that each poses different implications for the entity(ies) responsible for supporting resilience on the grid:

• Enhanced Status Quo. In this model, utilities will continue to manage their generation and/or delivery infrastructure to supply power to customers as today. At the same time, utilities will continue to invest in replacing aging infrastructure and advanced grid technologies to improve system reliability and resilience under traditional regulatory cost-of-service, ratemaking, and cost-recovery models (including revenue decoupling, in which utility cost recovery is delinked from volumetric electricity sales).
• Network Service Provider. As a more distributed energy future unfolds, the distribution system becomes a platform for enabling DERs to provide services to the wholesale market and as “non-wires alternatives” (so called because targeted installation of DERs can defer the need for transmission expansion). This model expands the role and value of the distribution system. This is accomplished by providing open access distribution services enabled by advanced technologies to allow the integration of high levels of DERs. Distribution services are based on network access fees comprised of demand charge and fixed charge components. Financial incentives for operational performance (e.g., for reliability and interconnections) and earnings mechanisms on DER non-wires grid services are employed. Otherwise, the traditional regulatory and utility economic model remains.
• Market Enabler. This model focuses on expanding the role of the utility distribution operations to become the distribution system (or market) operator (DSO). This “total DSO” (De Martini and Kristov, 2015) has responsibility for balancing demand and supply as well as distribution network reliability for a distribution area to an interchange point with the bulk power system operator. In this role, the DSO provides a single aggregated interface with the ISO/RTO, requiring the DSO to optimally dispatch DERs within its area. Traditional regulatory and utility economic models apply, along with the incentives above and market-based pricing for optional competitive services.
• Solutions Integrator. This model focuses on developing customer DER assets alongside other energy services, such as power and natural gas commodity supply, energy information services, and energy efficiency retrofits. In this model, utilities provide turn-key or selected engineering, procurement and construction services to support reliability, enhancement projects, customer high-voltage infrastructure, microgrid, and DER implementation. Services may also include customized engineering and operational consulting as well as emissions management and equipment condition assessment to ensure safety and reliability.

A critical factor in the transitions of the electricity sector is that continuing reductions in the cost and accelerating deployment of DERs is leading to a new class of customer that is both an electricity consumer and producer (“Prosumer”). There are now large and small prosumers who are increasingly interested in managing various aspects of their own electricity usage and supply. This is also enabling greater customer choice for installing select DER technologies to satisfy individual customer requirements associated with reliability, redundancy, and power quality. Whereas most backup power requirements in the past relied on diesel generators, numerous other DER technologies can supplant

or even replace the diesel generator as a backup power option. However, DERs have complex impacts on resilience, which are discussed in the following sections and throughout the report.

Distributed Energy Resources and the Distribution and Transmission Systems

DERs can provide benefits not only to the customers that employ them directly, but also to the broader transmission and distribution system. For example, DERs may help avoid or defer the need for new generation, transmission, or distribution infrastructure to address congestion, localized reliability, or resilience issues. The value of DERs for reliability, efficiency, and resilience depends upon their location and their particular attributes (e.g., their durability, their ability to be controlled, their availability when needed, the times of day when they reduce net load to the grid). Absent effective planning, DERs can also impose costs on the electricity system—for example, through equipment upgrades necessary to handle generation on distribution circuits, sub-optimal DER placement that contributes to congestion as opposed to alleviating it, and incomplete or inefficient sharing of information across the distribution-transmission interface.

This is particularly true at the distribution-system level, but also for interactions with the transmission grid. On the planning side, DERs can interact with the transmission system in several ways. First, behind-the-meter DERs complicate regional load forecasting, the process used to predict customer electricity demand at least 10 years into the future. Transmission system planners design the high-voltage system to meet forecasted demand. DERs behind the meter that provide energy to their owners have the potential to decrease load forecasts by the local retail utilities, which may account for DERs in their forecasting. Bulk power system planners may not be aware of DERs, and their load forecasts may not reflect the locations and types of DERs appearing or expected to appear on the system (NERC, 2016b).²⁰

DERs can also be used in transmission-system planning processes to address specific system needs identified through modeling that informs planning. If a planned generating unit retirement or predicted demand increase may lead to a localized reliability issue, DERs could be employed to address that issue in lieu of a more traditional solution like a substation upgrade or new transmission line. Several legal, operational, and institutional barriers to employing DERs as transmission-system solutions exist, but the potential is real.²¹ The use of DERs to address transmission-system limitations may also increase resilience in that the resources are more readily available after an outage or disturbance that could knock out a substation or transmission line for significant periods of time.

On the market design and operations side, DERs also have implications for the transmission system. In addition to potentially reducing the capacity-procurement needs of a region, DERs are legally able to participate in wholesale energy, capacity, and ancillary service markets. These centralized markets exist only within the RTO and ISO regions shown in Figure 2.5; the rest of the transmission-owning utilities rely on bilateral contracting or self-supply to meet their electricity needs.²² Some DERs have made progress in wholesale market participation. In PJM, for example, demand response resources participating in the wholesale market totaled more than 9,800 MW, with resources positioned at more than 17,000 locations across the PJM footprint (McAnany, 2017).

On both the transmission planning and wholesale market sides, a lack of operational awareness and coordination between distribution utilities (or, in the future, “distribution system operators”) and transmission-owning utilities, or the RTOs or ISOs operating the transmission system and wholesale power dispatch, serve as additional barriers to capturing the full potential value of DERs to the electric system. DER owners must understand what planning and market opportunities exist at both the distribution and transmission levels, and utilities and market operators must understand when resources are available for their use and when they are otherwise committed to provide grid services that render them unavailable for other uses.

Finding: The value of DERs for reliability, efficiency, and resilience depends upon their location, their attributes, the planning process behind their installation, and the legal and regulatory environment in which they are operated. While they can contribute to reliability and resilience, absent effective planning and an appropriate regulatory environment, DERs can also impose vulnerabilities and costs on the distribution system.

Other Technology Developments

Other new and emerging technologies may have important impacts on the structure and operation of the power system, including lower cost batteries as well as falling cost and growing capabilities of power electronics. Energy storage in the distribution system and on the customer side of the meter is a relatively new phenomenon. Some distributed energy storage (DES) is provided by thermal systems such as

___________________

²⁰ For example, the RTO that covers 13 Mid-Atlantic states and the District of Columbia, called PJM, was able to decrease its load forecast by 6,000 MW for 2020 by incorporating the energy efficiency and distributed solar that exists or is planned to come online between now and then (PJM, 2016).

²¹ See Southern California Edison and Consolidated Edison projects discussed in Tierney (2016b).

²² One notable exception is the recent development of an Energy Imbalance Market (EIM) administered by the CAISO, with participation by a growing number of utility systems in the Western grid. As of 2017, several electric utilities in Arizona, California, Idaho, Nevada, Oregon, Utah, and Wyoming had joined or are planning to join the EIM (CAISO, 2017).

hot water heaters. Other DES technologies involve chemical (e.g., battery) solutions. There is large variation in projected battery costs, potentially declining from today’s levels of about $600/kWh for whole battery systems to the range of $200–$300/kWh by the early 2020s. Lower cost batteries are providing interesting opportunities. Customers are installing on-site battery systems behind the meter in service areas with high charges for peak power consumption to shift their usage to off-peak periods. In general, energy storage has the potential to enable the electric system to become more efficient while enabling customer-side energy management (Navigant Research, 2013).

Over the next 20 years, customers will likely have greater technological opportunities to go entirely off grid, satisfying their electricity requirements with a combination of on-site generation and storage technologies. Customers capable of investing in such packages of technologies (or purchasing such services from the utility or a third party) may be able to take personal responsibility for their own resilient electric service. Although the committee believes the share of total customers taking advantage of such approaches will be limited, trends in grid defection and the technologies that could enable it should be closely monitored. Broader impacts on social equity will also warrant attention.

The controllability of DERs is enabled by low-cost computing and communications technologies. The internet of things and edge computing have progressed to the point where the capability to control DERs at low cost has become much more practicable, with significant advances even over the past few years. There is also significant experience among a number of utilities and third-party aggregators implementing and operating “smart grid” technologies that include operation of distributed generation, storage, and demand response. Fundamentally, the computing and communications technologies are not the limiting factor for adopting these control strategies, although they will require increasing sophistication and resolution in the monitoring and control systems used at the individual feeder and substation scale to understand and optimize circuit health and behavior.

Most organizations that have employed various DER strategies on a large scale have discovered that the need for “big data” analytics and other strategies to optimize the operation and control of these distributed assets is nascent, and more effort is needed to further develop the algorithms to enhance system operations and resilience by managing DER deployment. This is particularly true during off-normal conditions where the DER might be providing emergency backup power to support system restoration. Finally, these DER assets will necessarily need to interact with each other seamlessly, including during normal and off-normal or emergency situations, and not create or exacerbate any adverse conditions. These include but are not limited to hazards to utility workers and the public, equipment damage, and suboptimal operation of the remaining electrical assets.

Interdependencies Between the Electric and Natural Gas Infrastructure

One outcome of the trends under way in the electric system is the industry’s overall reliance on natural gas to fuel power generation, which increases the electric system’s reliability on conditions in the gas industry. This has potential implications for the resilience of the grid. The conventional wisdom is that the electric industry will become even more dependent upon natural gas than it has in recent years, and the natural gas industry looks to a future in which significant growth in demand depends upon developments in the power sector. For the electric system to become more reliable and resilient, attention must be paid to assure robust systems and practices across the two industries.

For many years, these two systems developed on largely different paths, from physical, economic, engineering, institutional, industrial-organizational, and regulatory perspectives. Both industries evolved with some degree of vertical integration and with aspects of each industry’s value chain regulated as monopolies by federal and/or state governments. The interconnected networks of each industry expanded over larger and larger geographic footprints. Recently, both systems have undergone eras of significant industry restructuring, with new players emerging as functions became unbundled and as competition entered into different parts of the business.

Today, however, each industry has its own set of cost structures, operating protocols and standards, commercial instruments, and pricing arrangements. Further, while the electric system operates as a network, following laws of physics on an interconnected grid rather than ownership or contract paths, the natural gas system is not a network industry. Individual companies own segments of the pipeline system, and users contract for access to and use of specific facilities. These changes also have occurred in parallel with dynamic developments in real-time, internet-based communications systems, complicating the interdependencies and allowing opportunities for new arrangements and solutions.

Today, natural gas supply still tends to move long distances from production sources to users’ sites, typically to locations where there is little to no storage close to or on the end-user’s property. This means that from an operational point of view, gas resources need to move “just in time” (i.e., they are used as they are delivered) to the end user through pipelines. During certain seasons and times of the day, many of these pathways—for example, those serving the Mid-Atlantic and Northeast regions—can become quite congested with firm gas deliveries, recognizing that gas injections at the production locations are intended to balance withdrawals of gas from the delivery system while taking in to account a variety of operational issues along the pathway from production to use. (“Just in time” delivery, however, sits within a context in which natural gas moves between 15–20 miles per hour on the interstate pipeline system, while

electric system operations occur at the sub-minute and multi-minute time frame.) Further, the growth in the power sector’s use of natural gas has not been accompanied in all relevant regions by expansions in pipeline capacity or increases in the efficiency of existing gas delivery infrastructure. Without change in some of the key features in current business models for competitive generators or in market rules, that situation is not expected to change dramatically in the near term, making it difficult to drive investment in pipeline/storage infrastructure based on demand from the electricity sector. (In some regions such as New England, however, changes in market rules have led many gas-fired generators to invest in dual-fuel [oil/natural gas] capability with on-site storage of oil as a lower-cost means to assure the ability to operate during periods when delivery of natural gas over pipelines is otherwise constrained.)

Regulatory issues at the intersection of gas and electric markets are complicated. While FERC may have responsibility for a broad set of policy issues on electric/gas integration issues, and NERC is evaluating the interdependencies from an operational and planning perspective, the states have strong interests and, in some cases, regulatory responsibilities that can affect market participants’ behaviors as well. Importantly, the structure of the natural gas production and delivery system in the United States does not have the same reliability requirements as now exist in the electric industry, and parts of that supply chain (e.g., production of natural gas) are effectively outside of FERC’s regulatory jurisdiction.

The electric and gas systems are already experiencing strains at their intersection. To date, integration issues related to increased gas-fired generation have caused rotating power outages in the Electric Reliability Council of Texas during the big freeze of 2011. And, owing to winter gas shortages and extreme cold weather, natural gas was either unavailable or priced too high for generators in PJM and the New York ISO during the polar vortex of 2014 (see Box 4.2 for a description of these events). In some regions, for example, generators need to commit to move gas volumes before knowing whether their offers into the RTO’s daily power markets have been accepted; conversely, generators need to offer prices into such energy markets without fully knowing the price and/or availability of their natural gas. There are other instances where gas customers that have contracted for firm gas supply and transportation service face potential (or real) curtailments as operational conditions change upstream and downstream. Tensions are visible across the business models of different players in the two industries and in the market rules in different regions. Further, there are different attitudes across the two industries regarding the urgency of anticipated changes in natural gas supply associated with growing use for electricity generation—specifically, the need for increased total supply and for that supply to be more nimble. It is difficult enough to introduce change into a single industry, where there may be players who perceive themselves as winning or losing from different options for resolving small and large issues. It will undoubtedly be even more difficult to introduce sensible but meaningful changes affecting market participants in two industries.

Decisions by myriad market actors and institutions do not typically reflect coordinated information about the performance of systems either across industry segments (e.g., across the electric and gas industries) or within industry supply chains (e.g., from production sources across interstate transmission systems). In the context of the events that occur in one or more parts of the industries’ systems, this absence of coordination mechanism may make some aspects of resilience—preparing for outages so as to limit their impact, sustaining service during an outage, and/or in restoring the systems to normal operations after the event—difficult to realize.

Finding: The electric industry has become highly dependent upon natural gas, and the natural gas industry looks to a future in which significant growth in demand depends upon developments in the electricity sector. For the electric system to become more reliable and resilient, attention must be paid to assuring the availability of adequate natural gas resources at all periods of time, including through investment in natural gas infrastructure (e.g., contractual arrangements and siting and construction of pipelines or storage), where it is economical to do so, fuel diversity for electric generators and natural gas compressors, and the alignment of planning and operating practices across the two industries.

Emerging Electric Grid Jurisdictional Challenges

Historically, and despite the state-to-state and regional variations in grid regulation around the country, FERC, the states, and regulated utilities have operated within relatively clear jurisdictional boundaries. In an electric grid consisting predominantly of large and dispatchable central station power plants, it was clear that FERC had jurisdiction over wholesale electricity rates and interstate transmission, whereas states had regulatory authority over retail sales and delivery over local transmission and distribution systems into our homes, businesses, and industrial facilities. Power on the system generally flowed in one direction, from the generator all the way to the end-use customers.

Over the past decade, however, the increasing penetrations of DERs and smart grid technology that are relevant for resilience have begun to change the very way the grid operates (see Figures 2.1 and 2.9). The grid is increasingly an interconnected web rather than a straightforward series of one-way pathways. However, the federal, state, and other legal constructs dictating the role of DERs on distribution and transmission systems are in active review by FERC and states in the relevant regions. Although this is a constructive response, there remain many jurisdictional ambiguities, policy mismatches, and an inability to maximize the potential value of technological change toward grid reliability and

resilience. The emerging relationships between DERs and the transmission and distribution systems have greatly outpaced the laws and regulations that govern their interactions. The 80-year-old FPA never contemplated the modern and complex system that exists today. As a result, the relatively clear boundary between state and federal authority over the electric system has blurred to some extent, causing uncertainty, if not confusion, among policy makers and energy industry participants. Recent legal challenges taken up to the Supreme Court have begun to sort through aspects of unresolved jurisdictional questions, but several questions remain.²³

Jurisdictional issues are also emerging within the distribution and transmission systems themselves. On the distribution system side, regulations typically assume one-directional power flow and fail to contemplate most DERs, including microgrids. From a resilience perspective, microgrids are a particularly interesting development—but they are not without legal uncertainties. Most state regulations obligate utilities to provide distribution service to all customers within their territories. With that obligation often comes the right to be the exclusive distribution provider. Microgrids that would connect buildings or a broader area technically involve their own distribution service and so, in many cases, are prohibited by existing utility regulations.

On the transmission system, the FPA itself remains a barrier to increased DER participation. For example, in the regional system planning processes, the FPA allows for transmission owners to allocate and recover the costs of new transmission investment except for non-wires alternatives, which includes DERs that are traditionally regulated by the states. As noted, the relationship among emerging technologies, evolving business models, and outdated laws and regulations that dictate authority over electric grid activities are stressed by the rapidly changing composition of resources and services involved with the delivery of energy, resulting in significant uncertainty. This, in turn, creates challenges for resilience planning.

Finding: Any new local, state, or federal programs, regulations, or laws designed to increase grid resilience will have to navigate a labyrinth of existing state and federal laws (some of which are out of date) that shape the incentives (or disincentives) for undertaking investments and actions aimed at enhancing resilience. This creates challenges for resilience planning, especially in light of the essential role of electricity in providing critical services and powering the economy.

LONGER-TERM DRIVERS OF CHANGE AND ASSOCIATED CHALLENGES AND OPPORTUNITIES FOR RESILIENCE

There is, of course, no way to reliably predict what the power system will look like in 30 to 50 years. However, it is possible to identify a variety of developments that could shape that future and then seek strategies that will be robust across that range of possibilities. To that end, here the committee identifies and discusses a variety of factors that might shape the future evolution of the system. Planning for grid resilience needs to take into account the expectation that the grid and its various institutions, technological features, legal structure, and economics will change—and in ways unknown today.

The Nature and Scope of the Future Regulatory Environment

Recent years have witnessed a dramatic shift in the structure and regulatory environment in which the high-voltage transmission system operates. A similar transformation has not yet occurred at the level of the distribution system. Whether such a transformation will occur, and what form it might take, will likely have profound effects on the future evolution of the system. Will federal authority be expanded to include a larger role at the level of the distribution system (Figure 2.10), as could occur, for example, where customers with on-site generation sell surplus back into the grid and thus set up the possibility of federal jurisdiction where such injections of power were considered sales for resale? Many states would likely oppose such an expansion, in a continuing tension between state and federal oversight seen in previous legislation including various provisions of PURPA and EPAct 2005.²⁴ The latter specifies the following:

___________________

²³ These recent cases have clarified a few different jurisdictional principles: First, one Supreme Court decision called EPSA v. FERC determined that FERC has the authority to regulate DER participation in wholesale markets. This authority means that, under certain circumstances, states and the federal government will both have the ability to regulate DERs in the performance of different activities. Second, another high court decision (known as Hughes v. Talen Energy Marketing, LLC) recognized that states have the authority to engage in their own preferred resource procurement efforts, but that they cannot cross a line that would invade FERC’s exclusive authority to set wholesale energy rates. The Hughes decision has fewer direct implications for DERs that may be procured for resilience purposes than it does for supply-side generating resources like wind, solar, or natural gas power plants, but it is nonetheless important to keep in mind in resilience program design. Third, a Supreme Court case called Oneok v. Learjet, considering the Natural Gas Act, emphasized that the ability of the federal government to regulate one particular area does not necessarily preclude state regulation in the same area. Other challenges around the ability of states and the federal government to regulate certain aspects of grid activities that have implications for DERs are working their way through federal courts. Although the mentioned cases have provided certainty in some respects, a general climate of uncertainty exists in states’ attempts to design new DER-centered regulations and programs.

²⁴ For example, PURPA’s Sections 1251, 1252, and 1254, and section 1254 of EPAct 2005.

While the legal justification under which federal jurisdiction might be further expanded is unclear, there is certainly a possibility that such justification might evolve over time.

There is of course also the possibility that in some domains, local, state, or even regional regulatory responsibilities might be expanded. If larger differences develop among regulatory structures in different parts of the country, this could present a variety of complications. As pressure grows to adopt more innovative strategies to address resilience issues that impact large areas of interconnected systems, states and regions may decide they need to adopt more innovative approaches.

The possibility of greater grid defection by customers may result in those customers providing their own electricity, entirely removed from federal and state rate jurisdiction altogether. It is likely that this would occur only in situations where the customer disconnects entirely from the grid. In such instances, states may have to address the terms and conditions under which customers may exit from or reenter the local distribution to assure (among other things) that legacy costs associated with utilities’ planning to provide service to those customers are addressed, according to traditional cost-incurrence and equity principles of utility regulation.

Penetration and Characteristics of Distributed Energy Resources

Closely linked to the way in which the future regulatory environment might evolve is the degree of penetration of distributed resources (Figure 2.11). The pace and extent of further deployment of DERs is the subject of major discussion in the industry. If the DOE SunShot targets are met, for example, rooftop solar will likely become cost competitive across much of the country without significant subsidies (Hagerman et al., 2016). Penetration of CHP has been much slower. Its future will depend in part on how the policy environment evolves and the wholesale-to-retail markup of natural gas. Costs are falling for local storage technology, but it is still only commercially viable in niche applications. Adoption could accelerate if costs fall and suppliers begin to offer storage with photovoltaic systems—with inverters and local intellegent control that reduces electricity bills and allows customers to continue to operate when grid power is unavailable.

There has been considerable discussion of smart controls for end-use devices, including the idea of “prices to devices” that would allow larger customers to decide when they will and will not operate particular electricity-using equipment given time-of-use pricing. While very extensive intelligent control is possible, what is less clear is when and whether the added hardware and intelligence will make economic sense.

Legal Implementation of Non-Utility Microgrids

Today in most of the United States, state law grants exclusive service territories to legacy distribution utilities, although there are a few exceptions.²⁵ This means that with the exception of a customer selling power back to the local utility, only that utility can distribute power to another entity. It also means that only a traditional utility can move power across a public road or other public right-of-way. If state laws were changed in such a way as to allow small-scale microgrids (larger than a few MWs) to be operated by private entities—with tariffs that symmetrically recognize the contributions of DERs while keeping the distribution company whole—the adoption of DERs could accelerate. Utility executives often argue that such a change would impose serious operational problems. However, from a technical point of view, there is very little difference between the two situations shown in Figure 2.12.

The committee asked several state regulatory agencies whether, in their jurisdictions, an entity other than the local distribution utility could build a small microgrid (e.g., less than a few 10s of MW), sell electric power to other entities, and be interconnected to the distribution utility. Several states noted that, as a matter of law, this was simply impossible in their states. Others indicated that the answer was more complex—an entity that wanted to engage in such activity would need to become a licensed and regulated utility. For example, staff of the Pennsylvania PUC said, “It is conceivable that an entity could perform such a function if they were properly licensed by the commission and the RTO and PJM. There may be some other legal factors that could limit their ability to sell power to entities other than the distribution utility and/or PJM Pennsylvania does allow net metering (see footnote 21) up to 3 MW.” Staff from the ICC noted, “Third parties that sell electric power to retail customers of an investor-owned utility must be licensed by the (ICC).” Staff of the New Hampshire Commission noted that in addition to having net metering, their state also has “group net metering (up to 1 MW).”

___________________

²¹ See Southern California Edison and Consolidated Edison projects discussed in Tierney (2016b).

²⁵ New York is one exception where the state may grant multiple franchises to serve a particular location; however, it is then up to local municipalities to grant easements along public streets and roads in order for the utility to install necessary facilities. Some Pennsylvania communities have been granted multiple franchises resulting in different utilities’ distribution lines on opposite sides of the street with service drops to customers crossing overhead. Nonetheless, in most regions service franchises are granted exclusively to one provider.

For years, the regulatory framing under which electric power has been provided in the United States was built on a foundation of universal service—that is, that access to basic electric power is to some degree a right that all citizens should enjoy. Indeed, it was this belief that prompted the creation of the Rural Electrification Administration in 1935 to supply power across rural America to customers whose locations were too remote to be attractive to privately operated utilities.

Today, the technical capability exists to provide different levels of service to different customers. This raises policy questions about whether all customers deserve some basic level of reliable service on the grounds of equity. As discussed in Chapter 5 of this report, there are ways in which distribution systems that contain advanced automation and distributed generation could be “islanded” so as to provide some limited service in the event of a large-area, long-duration blackout of the bulk power system. How the incremental cost of such upgrades should be covered, and whether they should only be based on an end-use customer’s ability to pay, raises obvious issues of social equity.

Over time, there will likely be greater opportunities for customers to defect from the grid (i.e., provide all of their electricity needs with customer-owned generation and storage). The goal of ensuring that all customers have access to electricity service that is affordable and reliable, combined with society’s larger interest in assuring that a resilient electric system supports the availability of critical social services, suggests that policy makers should continue to pay close attention to this trend. Policy makers may need to pursue mechanisms that encourage grid integration as part of service and to ensure that grid defection does not adversely impact those customers who have no practical economic choice but to remain dependent on the electric system to serve their needs.

Impacts of a Changing Climate

There remains uncertainty regarding how climate change and associated concerns will impact the electric power system (Figure 2.13). While the impacts of climate change will unfold over the coming decades, policy choices made in the near future can have a profound impact on the extent of that change (White House, 2016). The changing climate will result in more frequent and more intense extreme events (Melillo et al., 2014) that will impose damage and other challenges on the power system. Higher ambient temperatures will create increased demand for system cooling. In some parts of the country, it will also bring deeper and more prolonged droughts that, in turn, will result in problems of securing sufficient water for system cooling unless traditional wet cooling is replaced with dry cooling. In some locations, such as coastal regions prone to rising sea levels and storm surge or inland locations prone to frequent wildfires or flooding, it may prove necessary to relocate some facilities. Climate change will likely also result in new demands for electric power including larger air conditioning loads and, in some locations, an increased demand for power to pump water.

Changes in the Sources of Bulk Power

The past few decades have seen dramatic shifts in the sources of bulk power employed in the United States, and uncertainty persists regarding the future (Figure 2.14). Natural gas has displaced generation at many coal-fired baseload power plants, and even existing nuclear plants are retiring before the end of their operating licenses. However, if prices once again become higher or more volatile, investors may shy away from putting capital into natural gas plants and the trend could be reversed, as it was in the past.

Many observers anticipate significant penetration of new renewables, especially wind, solar, and hydro power. Today, wind generation constitutes approximately 5 percent of total U.S. generation, but a number of analyses suggest that there is no technical reason why the nation could not generate more than 60 percent of its electricity from wind. However, achieving such a high level of penetration would impose considerable requirements on land use, both for siting the wind turbines and for constructing the necessary transmission infrastructure, much of which will need to cross state

lines (MacDonald et al., 2016). Hence there is considerable uncertainty about the degree of future penetration of wind generation. Similar observations have been made with respect to solar generation. Many have argued that extensive use of biomass fuel, perhaps also with carbon capture and sequestration, will be necessary to achieve the objective of holding global warming to ≤ 2°C. At the same time, the widespread use of biomass imposes considerable logistical requirements and demands on land use (LaTourrette et al., 2011). Hence, it remains unclear how much future development will occur.

Nuclear power has contributed roughly 20 percent of the nation’s electricity generation for the past few decades. Many forecasts of U.S. energy production continue to assume their continued contribution of roughly the same share of supply. With the cost pressures that nuclear plants are facing from inexpensive natural gas and subsidized renewables, and uncertainties about the cost and likelihood of life extension and relicensing, a number of plants have closed recently. New York state and Illinois recently adopted policies designed to keep existing plants operating (McGeehan, 2016). The only new plants under construction in the United States are in the service territory of vertically integrated utilities in the Southeast, where costs can be included in the rate base. In addition, the nation has largely abandoned aggressive research on more advanced reactor designs, so that for at least the next several decades the only options for new nuclear construction will likely be existing light-water reactor designs (DOE, 2017c; Ford et al., 2017). There may be some renewed interest in advanced reactor design research (DOE, 2017c), but the extent of programmatic support for this vision remains uncertain. Small modular reactors have received a lot of attention in part because they require less capital investment and offer much greater siting flexibility. Despite these benefits, however, long-standing efforts have never reached commercial construction (Larson, 2016). Investment in new, small, and advanced reactors may require a number of changes in business models and reactor designs that allow standardized and quicker manufacturing of components and construction of reactors.

Today, technologies for cost-effective bulk storage are limited. Pumped hydro storage imposes considerable land use and other environmental costs, and only a few facilities for compressed air storage have been built. Battery storage is beginning to have some impact on the power system, especially in behind-the-meter applications. In 2012, DOE established the Joint Center for Energy Storage Research (JCESR) as one of its “Energy Innovation Hubs.” JCESR’s stated goal is to “deliver electrical energy storage with five times the energy density and one-fifth the cost” of present storage technologies (Crabtree, 2016). In addition to striving to develop batteries that would allow all electric passenger vehicles to be profitably marketed at a cost of approximately $20,000 and with a range of 200 miles, JCESR director George Crabtree has articulated remarkably aggressive goals for affordable grid storage, including battery technology that would be competitive with pumped hydro storage, chemically based, and capable of seasonal storage. However, battery experts with whom the committee discussed the JCESR goals for bulk grid storage have expressed considerable doubt about achieving those goals, especially on the time scale of the next several decades.

Nonetheless, all electric vehicles with those capabilities would have an impact on both the transportation sector and on electricity demand. Whether or not the JCESR goals are met, a much higher penetration of electric or hybrid vehicles may well occur on the time scale of the next several decades. With greater adoption of electric and plug-in hybrid vehicles, there may be greater opportunities for using connected vehicle batteries to improve grid resilience—for example,

by using electric vehicle batteries to provide a fraction of a home’s electricity demand during a large-area, long-duration outage (see Chapter 5).

SUSTAINING AND IMPROVING THE RESILIENCE OF A GRID THAT IS CHANGING RAPIDLY AND IN UNCERTAIN WAYS

From all of the foregoing, five things are apparent:

1. The grid is undergoing dramatic change. This will be especially true over the next few years at the distribution level where DERs continue to increase and change the relationship of utilities to end users. While DERs may provide many opportunities to increase grid resilience, this will require regulatory changes and effective planning and coordination. Over the next decade or two, major changes are also likely in bulk power transmission.
2. Much of the hardware that makes up the grid is long lived, which limits the rate of change in the industry. However, over periods of a decade or two, many changes are possible, and it is virtually impossible to know how the future grid will evolve.
3. No single entity is in charge of planning the evolution of the grid. That will become ever more true as more and more players become involved, particularly regarding deployment and operation of DERs at the distribution level.
4. All players will be concerned about reliability, both for themselves and collectively. Only a few are likely to be focused in a serious way on identifying growing system-wide vulnerabilities or identifying changes needed to assure resilience.
5. Today, virtually no one has a primary mission of building and sustaining increased system-wide resilience or developing strategies to cover the cost of investments to increase resilience in the face of low probability events that could have very large economic and broader social consequences.

These five observations carry profound implications for the future resilience of the power system. In Chapter 3, the committee explores the many types of events that can give rise to large-area, long-duration outages. Chapters 4, 5, and 6 correspond to the three stages of the resilience framework illustrated in Figure 1.2, making specific recommendations in the course of the discussion. Finally, in Chapter 7 the committee both summarizes those recommendations and comes back to the broader implications of the five observations above to consider an integrated perspective to the issue of electricity system resilience and how best to assure that continued attention is directed at building and sustaining system-wide resilience of the nation’s power system.

REFERENCES

DSIRE (Database of State Incentives for Renewables and Efficiency). 2016a. “Renewable Portfolio Standard Policies.” http://www.dsireusa.org/resources/detailed-summary-maps/. Accessed July 13, 2017.

DSIRE. 2016b. “3rd Party Solar PV Power Purchase Agreement (PPA).” http://ncsolarcen-prod.s3.amazonaws.com/wp-content/uploads/2014/11/3rd-Party-PPA.pdf.

DSIRE. 2016c. “Net Metering.” http://ncsolarcen-prod.s3.amazonaws.com/wp-content/uploads/2016/07/Net_Metering1.pdf.

DSIRE. 2016d. “Find Policies & Incentives by State.” http://www.dsireusa.org/. Accessed September 22, 2016.

EEI (Edison Electric Institute). 2017. “Industry Capital Expenditures.” http://www.eei.org/resourcesandmedia/industrydataanalysis/industryfinancialanalysis/QtrlyFinancialUpdates/Documents/EEI_Industry_Ca-pex_Functional_2017.04.21.pptx. Accessed July 13, 2017.

EIA (Energy Information Administration). 2010. “Electricity: Status of Electricity Restructuring by State.” http://www.eia.gov/electricity/policies/restructuring/restructure_elect.html. Accessed September 22, 2016.

EIA. 2015. “What is U.S. Electricity by Generation Source?” https://www.eia.gov/tools/faqs/faq.cfm?id=427&t=3. Accessed September 22, 2016.

EIA. 2016a. Electric Power Sales, Revenue, and Energy Efficiency. Form EIA-861, Detailed Data Files. https://www.eia.gov/electricity/data/eia861/. Accessed July 13, 2017.

EIA. 2016b. Monthly Energy Review. https://www.eia.gov/totalenergy/data/monthly/pdf/mer.pdf.

EIA. 2016c. “Preliminary Monthly Electric Generator Inventory.” http://www.eia.gov/electricity/data/eia860m/. Accessed July 13, 2017.

EIA. 2016d. “Today in Energy: Demand Trends, Prices, and Policies Drive Recent Electric Generation Capacity Additions.” http://www.eia.gov/todayinenergy/detail.cfm?id=25432. Accessed September 20, 2016.

EIA. 2016e. “Today in Energy: Solar, Natural Gas, Wind Make Up Most 2016 Generation Additions.” http://www.eia.gov/todayinenergy/detail.php?id=29212. Accessed December 19, 2016.

EPRI (Electric Power Research Institute). 2011. Needed: A Grid Operating System to Facilitate Grid Transformation. https://www.smartgrid.gov/files/Needed_Grid_Operating_System_to_Facilitate_Grid_Transformati_201108.pdf.

EPRI. 2015. The Integrated Grid: A Benefit-Cost Framework. Final Report, 3002004878. Palo Alto, Calif.: EPRI.

FERC (Federal Energy Regulatory Commission). 2016a. “Regional Transmission Organizations (RTO)/Independent System Operators (ISO).” https://www.ferc.gov/industries/electric/indus-act/rto.asp. Accessed July 13, 2017.

FERC. 2016b. Assessment of Demand Response and Advanced Metering. https://www.ferc.gov/legal/staff-reports/2016/DR-AM-Report2016.pdf.

Ford, M.J., A. Abdulla, M.G. Morgan, and D.G. Victor. 2017. Expert assessments of the state of U.S. advanced fission innovation. Energy Policy 108: 194–200.

Glass, J. 2016. “Enhancing the Resiliency of the Nation’s Electric Power Transmission and Distribution System,” presentation to the Committee on Enhancing the Resilience of the Nation’s Electric Power Transmission and Distribution System, September 29, Washington, D.C.

GMLC (Grid Modernization Laboratory Consortium). 2017. Grid Modernization: Metrics Analysis. Richland, Wash.: Pacific Northwest National Laboratory.

Gridwise Alliance. 2016. “3rd Annual Grid Modernization Index.” http://www.gridwise.org/report_download.asp?id=17. Accessed July 13, 2017.

Hagerman, S., P. Jaramillo, and M.G. Morgan. 2016. Is rooftop solar PV at socket parity without subsidies? Energy Policy 89: 84–94.

Hirschman, A. 1970. Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations and States. Cambridge: Harvard University Press.

Hoovers. 2017. “Arizona Public Service Company Revenue and Financial Data.” http://www.hoovers.com/company-information/cs/revenuefinancial.arizona_public_service_company.959a800ac6670f2a.html. Accessed February 10, 2017.

IEEE (Institute of Electrical and Electronics Engineers). 2013. Standard for Interconnecting Distributed Resources with Electric Power Systems. http://grouper.ieee.org/groups/scc21/1547/1547_index.html. Accessed July 13, 2017.

IRC (ISO/RTO Council). 2015. “Members at a Glance.” http://www.isorto.org/About/Members/allmembers. Accessed December 18, 2016.

Karlin, B. 2012. Public Acceptance of Smart Meters: Integrating Psychology and Practice. ACEEE Summer Study on Energy Efficiency in Buildings. http://aceee.org/files/proceedings/2012/data/papers/0193-000243.pdf.

Keogh, M., and C. Cody. 2013. Resilience in Regulated Utilities. National Association of Regulatory Utility Commissioners’ Grants and Research Department, November. https://pubs.naruc.org/pub/536F07E4-2354D714-5153-7A80198A436D. Accessed July 13, 2017.

Kwasinski, A. 2016. Quantitative model and metrics of electric grids’ resilience evaluated at a power distribution level. Energies 9(93).

Larson, A. 2016. “Is There a Market for Small Modular Reactors?” Power Magazine, June 1. http://www.powermag.com/market-small-modularreactors/. Accessed July 13, 2017.

LaTourrette, T., D.S. Ortiz, I. Hlavka, N. Burger, and G. Cecchine. 2011. Supplying Biomass to Power Plants: AModel of the Costs of Utilizing Agricultural Biomass in Cofired Power Plants. Santa Monica, Calif.: Rand Corporation. http://www.rand.org/content/dam/rand/pubs/technical_reports/2011/RAND_TR876.pdf.

Lave, L.B., J. Apt, and S. Blumsack. 2004. Rethinking electricity deregulation. The Electricity Journal 17(8): 11–26.

Lazard. 2015. Lazard’s Levelized Cost of Energy Analysis—Version 9.0. https://www.lazard.com/media/2390/lazards-levelized-cost-of-energyanalysis-90.pdf.

MacDonald, A.E., C.T. Clack, A. Alexander, A. Dunbar, J. Wilczak, and Y. Xie. 2016. Future cost-competitive electricity systems and their impact on US CO2 emissions. Nature Climate Change 6: 526–531.

McAnany, J. 2017. 2016 Demand Response Operations Market Activities Report. http://www.pjm.com/~/media/markets-ops/dsr/2016-demandresponse-activity-report.ashx. Accessed July 13, 2017.

McGeehan, P. 2016. “New York State Aiding Nuclear Plants with Millions in Subsidies” New York Times, August 1. http://www.nytimes.com/2016/08/02/nyregion/new-york-state-aiding-nuclear-plants-with-millions-in-subsidies.html?_r=0. Accessed January 2, 2017.

Melillo, J.M., T.C. Richmond, and G.W. Yohe. 2014. Climate Change Impacts in the United States: The Third National Climate Assessment. U.S. Global Change Research Program. doi:10.7930/J0Z31WJ2.

MIT (Massachusetts Institute of Technology). 2011. The Future of the Electric Grid. http://energy.mit.edu/wp-content/uploads/2011/12/MITEI-The-Future-of-the-Electric-Grid.pdf.

MIT. 2016. Utility of the Future: An MIT Energy Initiative Response to an Industry in Transition. http://energy.mit.edu/wp-content/uploads/2016/12/Utility-of-the-Future-Full-Report.pdf.

NAE (National Academy of Engineering). 2003. “Greatest Engineering Achievements of the 20th Century.” http://www.greatachievements.org/. Accessed September 22, 2016.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2016. The Power of Change: Innovation for Development and Deployment of Increasingly Clean Electric Power Technologies. Washington, D.C.: The National Academies Press.

Navigant Research. 2013. The Lithium Ion Battery Market. https://www.arpa-e.energy.gov/sites/default/files/documents/files/Jaffe_RANGE_Kickoff_2014.pdf.

NERC (North American Electric Reliability Corporation). 2010. “Functional Model.” http://www.nerc.com/pa/Stand/Pages/FunctionalModel.aspx. Accessed September 22, 2016.

NERC. 2011. Special Report: Spare Equipment Database System. Atlanta: NERC.

NERC. 2014. NERC CIP-014 Physical Security Standards. http://www.nerc.com/pa/Stand/Reliability%20Standards/CIP-014-1.pdf.

NERC. 2016a. Glossary of Terms Used in NERC Reliability Standards. http://www.nerc.com/files/glossary_of_terms.pdf.

NERC. 2016b. Distributed Energy Resources: Connection, Modeling and Reliability Considerations. http://www.nerc.com/comm/Other/essntlrlbltysrvcstskfrcDL/May%202016%20Meeting%20Materials.pdf.

NERC. 2017. “Critical Infrastructure Protection Compliance.” http://www.nerc.com/pa/CI/Comp/Pages/default.aspx. Access March 13, 2017.

NJBPU (New Jersey Board of Public Utilities). 2015. Record of Decision, May 21. http://www.state.nj.us/bpu/pdf/boardorders/2014/20140521/521-14-2I.pdf.

NJBPU. 2017. In the Matter of the Petition of Rockland Electric Company for Approval of an Advanced Metering Program; and for Other Relief. BPU Docket No. ER16060524. http://www.nj.gov/rpa/docs/ER16060524_Rate_Counsel_Initial_Brief_Rockland_AMI.pdf. Accessed July 13, 2017.

NRC (National Research Council). 2012. Terrorism and the Electric Power Delivery System. Washington, D.C.: The National Academies Press.

Parfomak, P.W. 2014. Physical Security of the U.S. Power Grid: High-Voltage Transformer Substations. https://fas.org/sgp/crs/homesec/R43604.pdf.

PJM. 2016. Load Forecast Report, January. http://www.pjm.com/~/media/documents/reports/2016-load-report.ashx. Accessed September 21, 2016.

PJM. 2017. “Fact at a Glance.” https://www.pjm.com/~/media/about-pjm/newsroom/fact-sheets/pjm-at-a-glance.ashx. Accessed February 20, 2017.

Platts. 2014. “Utility Service Territories of North America.” http://www.platts.com/products/utility-service-territories-north-america-map. Accessed July 13, 2017.

Plug-in America. 2016. “State and Federal Incentives.” https://pluginamerica.org/why-go-plug-in/state-federal-incentives/. Accessed September 21, 2016.

PNNL (Pacific Northwest National Laboratory). 2015. The Emerging Interdependence of the Electric Power Grid & Information and Communication Technology. PNNL-24643, August. http://www.pnnl.gov/main/publications/external/technical_reports/PNNL-24643.pdf.

Reuters. 2010. “Smart Grid Skepticism Derails Baltimore Plan.” http://blogs.reuters.com/great-debate/2010/06/23/smart-grid-scepticism-derails-baltimore-plan/. Accessed July 13, 2017.

RGGI (Regional Greenhouse Gas Initiative). 2016. “CO₂ Auctions, Tracking & Offsets.” https://www.rggi.org/market. Accessed September 21, 2016.

SNL (Sandia National Laboratories). 2014. Conceptual Framework for Developing Resilience Metrics for the Electricity, Oil, and Gas Sectors in the United States. https://cfwebprod.sandia.gov/cfdocs/CompResearch/docs/EnergyResilienceReportSAND2014-18019o.pdf.

SoCo (Southern Company). 2017. “Overview of Our Business.” http://www.southerncompany.com/about-us/our-business/home.cshtml#. Accessed February 20, 2017.

Spence, A., C. Demski, C. Butler, K. Parkhill, and N. Pidgeon. 2015. Public perceptions of demand-side management and a smarter energy future. Nature Climate Change 5: 550–554.

State of New York. 2014. Reforming the Energy Vision: NYS Department of Public Service Staff Report and Proposal. CASE 14-M-0101. http://www3.dps.ny.gov/W/PSCWeb.nsf/All/C12C0A18F55877E785257E6F005D533E?OpenDocument. Accessed September 21, 2016.

TCR (Tabors Caramanis Rudkevich). 2016. “Developing Competitive Electricity Markets and Pricing Structures.” White paper prepared for New York State Energy Research and Development Authority and New York State Department of Public Service, Contract 64271. https://www.hks.harvard.edu/hepg/Papers/2016/TCR.%20White%20Paper%20on%20Developing%20Competitive%20Electricity%20Markets%20and%20Pricing%20Structures.pdf.

Tierney, S. 2016a. The U.S. Coal Industry: Challenging Transitions in the 21st Century. http://www.analysisgroup.com/uploadedfiles/content/insights/publishing/tierney%20-%20coal%20industry%20-%2021st%20century%20challenges%209-26-2016.pdf.

Tierney, S. 2016b. The Value of “DER” to “D”: The Role of Distributed Energy Resources in Supporting Local Electric Distribution System Reliability. http://www.cpuc.ca.gov/uploadedFiles/CPUC_Public_Website/Content/About_Us/Organization/Divisions/Policy_and_Planning/Thought_Leaders_Events/Tierney%20White%20Paper%20%20Value%20of%20DER%20to%20D%20-%203-30-2016%20FI-NAL.pdf.

USCB (U.S. Census Bureau). 2016. “QuickFacts.” http://www.census.gov/quickfacts/. Accessed December 18, 2016.

USDA (U.S. Department of Agriculture). 2016. “Rural Utilities Service.” https://www.rd.usda.gov/about-rd/agencies/rural-utilities-service. Accessed July 13, 2017.

White House. 2016. Opportunities to Enhance the Nation’s Resilience to Climate Change. https://www.whitehouse.gov/sites/default/files/finalresilienceopportunitiesreport.pdf.

Willis, H.H., and K. Loa. 2015. Measuring the Resilience of Energy Distribution Systems. Santa Monica, Calif.: RAND Corporation.