2014 1201 APS 111(d)Comments CO2Stnd-ExistingEGUs FINAL … · EPA Docket No. EPA-HQ-OAR-2013-0602...

Chas Spell PO Box 53999 Telephone: 602-250-5340 Director Mail Station 9303 Fax: 602-250-3005 Environmental Policy & Programs Phoenix, AZ 85072-3999 e-mail: [email protected]

December 1, 2014 Electronically submitted Environmental Protection Agency EPA Docket Center (EPA/DC) Mail code 28221T 1200 Pennsylvania Ave. NW. Washington, DC 20460 Attention: Docket No. EPA-HQ-OAR-2013-0602 Re: APS Comments on EPA’s Proposed Rule Carbon Pollution Emissions Guidelines for Existing Stationary Sources: Electric Utility Generating Units To Whom It May Concern:

Arizona Public Service Company (“APS”) appreciates the opportunity to submit comments on

the U.S. Environmental Protection Agency’s (“EPA” or the “Agency”) proposed rule “Carbon

Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility Generating

Units” (the “Proposed Rule”) pursuant to section 111(d) of the Clean Air Act and the Notice of

Data Availability in support of the Proposed Rule.

As one of the Southwest’s largest electric utilities, APS operates in a manner that is protective of

the environment and is committed to environmental stewardship and sustainable operations.

APS operates a fleet of power generating facilities to serve the needs of over one million APS

customers; this fleet includes two coal-fired power plants, several natural gas-fired power plants,

and the nation’s largest nuclear power plant, Palo Verde Nuclear Generating Station. APS also

has an increasing portfolio of renewable energy, and manages a robust end-user energy

efficiency program. Thus, APS has a significant interest in the Proposed Rule.

EPA Docket No. EPA-HQ-OAR-2013-0602 December 1, 2014 Transmittal of APS Comments on Proposed Rule for Carbon Pollution Standards Under CAA §111(d) for Existing Stationary Sources

2

As discussed in the attached comments, APS has concerns with several different aspects of the

Proposed Rule. APS respectfully requests that EPA consider our comments in the development

of the final rule. Additionally, APS is an active member of the Utility Air Regulatory Group

(UARG), a voluntary, ad hoc, not-for-profit association of electric generating companies and

organizations and national trade associations. UARG submitted comments to this Proposed Rule

which APS fully supports and requests that EPA consider as well.1

Notwithstanding the significant legal concerns identified by APS in the enclosed comments, APS

wants to point out three specific issues associated with Building Block 2 along with proposed

solutions that would resolve most of the technical concerns identified in the comments. APS

worked in a collaborative effort with the other Arizona utilities, the Arizona Department of

Environmental Quality, and the Arizona Corporation Commission during the development of the

proposed solutions. EPA’s adoption of the recommended changes to the Building Block 2

methodology would result in a final carbon emission rate goal for Arizona of 963 lb MWh,

which is a 34 percent reduction from the baseline year. Furthermore, these proposed changes

will not have a material effect on a nationwide basis and will allow the achievement of the

carbon emission intensity reduction objectives of the Proposed Rule.

Issue #1: EPA should allow the states to set the interim goals

APS questions the actual benefits that will be achieved by interim carbon intensity goals;

however, at the very least, the states should be allowed to establish these goals. Allowing

the states to set the interim goals is more consistent with the legal requirements of the

Clean Air Act and because the states have a greater understanding of the specific state

issues, they are better suited to establish interim goals to place each state on a successful

path to achieve the final goal.

1 Comments of the Utility Air Regulatory Group on EPA’s Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility Generating Units; Proposed Rule, 79 Fed. Reg. 34,830, Docket ID No. EPA-HQ-OAR-2013-0602 (Dec. 1, 2014) [hereinafter “UARG Comments”].

EPA Docket No. EPA-HQ-OAR-2013-0602 December 1, 2014 Transmittal of APS Comments on Proposed Rule for Carbon Pollution Standards Under CAA §111(d) for Existing Stationary Sources

3

Issue #2: Establish provisions for remaining useful life

In the recently issued NODA, EPA is on the correct path for incorporating the remaining

useful life of affected units in the re-dispatch assessment. Such actions will greatly

reduce electric system reliability concerns and allow for more logical transition to

lowering the carbon intensity from existing electric generating units.

Issue #3: Adjust the natural gas emission rate for existing NGCC units

The existing NGCC units should not be held to a performance standard more stringent

that the standard for new NGCC units. To establish the carbon emission rate goals, EPA

used a single year emission rate from existing NGCC units. A single year emission rate

is not necessarily reflective of future operations for these units and existing NGCC units

should not be held to a performance standard more stringent than the performance

standard for new, efficient NGCC units.

APS appreciates the considerable effort that EPA has put forth in developing the Proposed Rule

and the extensive outreach by EPA to assure it develops the best possible rule concerning the

carbon intensity of existing sources. APS believes the enclosed comments, along with the

proposed solutions, will allow EPA to achieve this object. Once again, APS appreciates the

opportunity to provide comments.

Sincerely,

Chas Spell Director, Environmental Policy & Programs Arizona Public Service Company

COMMENTS OF ARIZONA PUBLIC SERVICE COMPANY ON EPA’S PROPOSED CARBON POLLUTION EMISSION GUIDELINES FOR EXISTING STATIONARY SOURCES:

ELECTRIC UTILITY GENERATING UNITS AND NOTICE OF DATA AVAILABILITY IN SUPPORT OF PROPOSED RULE

P a g e | 1

INTRODUCTION

APS fundamentally objects to the Environmental Protection Agency’s (“EPA” or “Agency”)

approach in the Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric

Utility Generating Units Proposed Rule (“Proposed Rule” or “Proposal”)2 on numerous grounds,

as articulated herein and in comments incorporated by reference. Clean Air Act (“CAA” or

“Act”) Section 111(d) dictates a framework of cooperative federalism in which EPA provides

emission guidelines to states and states develop and submit a plan establishing the performance

standards for existing sources. EPA’s Proposed Rule contravenes this principle of cooperative

federalism expressed throughout the CAA and its implementing regulations by establishing

mandatory differential state performance goals based on EPA’s determination of how and when

emission reductions are available in each state. In addition, EPA’s proposed rate-based carbon

dioxide (“carbon”) emission performance standards are calculated based on emission reduction

measures that the Agency itself does not have the authority to regulate. The performance

standards incorporate reductions that can only be achieved through energy system regulation.

EPA attempts to circumvent its lack of authority under federal law by inducing states to

implement these measures through state plans. Furthermore, EPA claims that its approach

affords states flexibility for compliance. But this flexibility rings hollow because EPA mandates

stringent performance standards that must be implemented on an expedited timeframe, based on

faulty assumptions, including that all measures can be carried out to the fullest extent in each

state in order to reach the proposed goals.

Furthermore, in the Proposed Rule EPA applies its newly developed best system of emission

reduction (“BSER”) to establish state-specific carbon emission rate interim and final goals. To

apply BSER, EPA had to make numerous technical assumptions. As a result of these

assumptions, Arizona is faced with serious issues regarding the reliability of its electrical energy

system and the associated financial impacts. 2 79 Fed. Reg. 34,830 (June 8, 2014).



P a g e | 2

I. EPA LACKS AUTHORITY TO IMPLEMENT A SYSTEM-BASED APPROACH AND MUST ESTABLISH A BEST SYSTEM OF EMISSION REDUCTION ACHIEVABLE AT THE INDIVIDUAL ELECTRIC GENERATING UNIT

A. EPA’s Authority to Establish the Best System of Emission Reduction Is Limited by the Plain Language of the Clean Air Act

EPA’s Section 111(d) authority to regulate is circumscribed under the CAA, which allows the

agency only to consider the individual electric generating unit (“EGU”), and not the entire

electric system or electricity consumers, when establishing the BSER.3 The plain language of

the statute makes clear that the BSER determination is limited to on-site controls, activities, or

work practices, and does not include beyond-the-fence measures that are outside of the purview

of CAA regulation and beyond the control of the regulated source.

First, the term “stationary source” in Section 111 is narrowly defined to “building[s],

structure[s], facilit[ies], or installation[s],” and does not include combinations or systems of

sources.4 Additionally, state plans are focused on particular sources: a state plan establishes

standards of performance for an “existing source,” and EPA “shall permit the State in applying a

standard of performance to any particular source under a [state] plan . . . to take into

consideration, among other factors, the remaining useful life of the existing source to which such

standard applies.”5

Section 111’s source-specific language makes clear Congress’s intent that BSER reflect what is

achievable at each individual affected source. When determining an achievable emission

reduction for a source under Section 111(d), EPA must first identify systems of emissions

reduction that have been adequately demonstrated for units within the source category, taking

into account the costs of achieving such reduction, non-air quality impacts, and energy

3 In addition, EPA lacks authority to regulate pollutants from any source category, including coal-fired EGUs, that is already regulated under CAA Section 112. APS incorporates UARG’s Comments, Section IV, on this point. 4 42 U.S.C. § 7411(a)(3). 5 42 U.S.C. § 7411(d)(1) (emphasis added).



P a g e | 3

considerations. Then, the Agency must identify the emission limits that are achievable through

the use of these systems of emissions reduction. Emissions performance and the ability to reduce

carbon emissions across the nation’s fossil fuel-fired fleet will vary according to fuel type, age,

combustion technology, operating conditions, and other unit-specific factors. Because what is

achievable at the individual source is fundamental to the determination of BSER and the

resulting standard of performance for an existing source, Section 111 does not authorize EPA to

consider emission reduction actions outside the fence line in its determinations.

B. EPA Has a Long History of Recognizing that Standards Must Be Achievable by the Affected Source

In addition to the language of the CAA itself, EPA’s Section 111 implementing regulations and

previous Section 111 rule-makings demonstrate that EPA has a long history of limiting the scope

of regulation to the affected source. For instance, under EPA’s general Section 111

implementing regulations, the Agency must develop specific emissions guidelines for states,

which must “reflect[] the application of the best system of emission reduction (considering the

cost of such reduction) that has been adequately demonstrated” for designated facilities.6 EPA

must specify different emission guidelines or compliance times or both for different sizes, types,

and classes of designated facilities to address the costs of control, physical or geographic

limitations, or similar factors.7 The continuous emphasis in Subpart B on “designated facilities”

clearly shows that EPA has always understood the scope of Section 111 to be limited to specific

regulated sources.8 EPA’s historical rule-making efforts underscore this understanding: every

Section 111 rule has been limited in scope to the regulated source.9

6 40 C.F.R. § 60.22(b)(5). See also 42 U.S.C. § 7411(a)(1). 7 40 C.F.R. § 60.22(b)(5). Although Section 111 affirmatively directs EPA to consider differences among classes, types, and size of sources within a category in establishing performance expectations, the Agency failed to do so in its Proposed Rule. 8 See also id. §§ 60.22(b)(3) (guideline documents will provide “information on the degree of emission reduction which is achievable with each system, together with information on the costs and environmental effects of applying each system to designated facilities” (emphasis added)); 60.21(e) (defining “emission guideline” as a guideline



P a g e | 4

C. EPA’s Proposed Rule Goes Far beyond EPA’s Regulatory Authority

EPA’s Proposed Rule turns the regulatory framework on its head. Under the clear language of

Section 111, EPA should first identify a source category and determine BSER based on what is

achievable by the affected source, and then leave it up to the states to set standards of

performance for affected sources, implementing BSER. However, EPA, in this rule-making, sets

mandatory emission performance standards based on measures outside of the affected sources’

control and leaves it to the state to identify which entities (EGUs or not) should be obligated

under the rule. This reverses the regulatory process and is clearly outside the scope of EPA’s

authority. Three of the four proposed building blocks on which EPA based its mandatory state

performance standards assume that reductions will come from beyond-the-fence entities over

which EPA lacks the authority to regulate. EPA’s attempt to regulate the state’s entire electric

power market is irreconcilable with the statutory language. Because the statute clearly requires

BSER to be achievable by a source, EPA cannot require reductions, such as the beyond-the-

fence measures, that would be beyond the source’s control.

D. Serious Enforceability Issues with EPA’s Proposed Rule, Stemming from Reliance on Beyond-the-Fence Measures, Demonstrate EPA’s Lack of Authority to Regulate

Serious enforceability problems inherent in EPA’s Proposed Rule demonstrate that Congress

cannot possibly have meant for the Agency to expand the scope of Section 111 as far as the

Proposed Rule. The Proposed Rule relies on reductions that cannot be achieved by affected

sources alone. EPA attempts to impose reduction obligations through building blocks 2, 3, and 4

that fall well outside of the scope of the CAA. Regulated EGUs cannot increase the share of

natural gas generation at units beyond their control, nor can they mandate an increase in

renewable energy generation or mandate measures requiring consumers to be energy efficient.

reflecting the “degree of emission reduction achievable through application of [BSER] which . . . has been adequately demonstrated for designated facilities” (emphasis added)). 9 APS refers to the UARG Comments, Section III.A.3 for an in-depth discussion of EPA’s historical rule-makings.



P a g e | 5

Neither does Section 111’s language, which specifically focuses on individual sources, allow

EPA to regulate any of these entities directly. And it does not authorize EPA to require states to

regulate any entities other than regulated sources. Yet, if a state in its plan includes measures

enforceable against non-source entities, EPA envisions that those measures become federally

enforceable by virtue of their inclusion in a state plan.10 This would increase the scope of EPA’s

authority under the CAA beyond recognition. Furthermore, if a state does not act in a timely

manner to provide an approvable plan that includes enforceable measures, then EPA is mandated

by the Act to create a federal plan. However, EPA would not have the authority to implement

any measure in Building Block 2, 3, or 4 as it cannot regulate, for instance, a state’s energy

consumers.

E. A Systems-Based Approach Is Fundamentally Flawed Because EPA Could Claim Limitless Authority in the Context of Other Sectors

The logical extension of EPA’s systems-based approach in the context of other sectors would

create near-boundless EPA authority. EPA has expanded its reach to the energy regulatory

system through its systems-based approach in the power sector. Applied to other sectors, such as

manufacturing or refining, EPA could define systems to encompass things, such as goods

movement, transportation fuels, and manufacturing processes. Such an approach would give

EPA extraordinary authority to favor certain lower-carbon modes of transportation, raw

materials, fuel choices, manufacturing processes, etc. There is no evidence that Congress

intended for EPA to dictate industrial or transportation (or, in this case, energy) policy to such a

degree. By forcing such a radical change from the nation’s current economic dispatch system of

electricity generation to one based on environmental dispatch, the Proposed Rule represents an

10 79 Fed. Reg. at 34,901.



P a g e | 6

overreach by the Agency into the energy markets. “EPA is systematically federalizing under the

Clean Air Act what was once in the clear purview of the states or the markets.”11

II. THE CLEAN AIR ACT DOES NOT ALLOW EPA TO TREAT SOURCES AS BOTH “NEW” AND “EXISTING”

In the Proposed Rule, EPA solicits comment on how new units should be treated under Section

111(d).12 APS opposes the incorporation of any new 111(b) sources into the 111(d) program.

The CAA is clear that EPA’s choice of regulations under Section 111 is binary: EPA may either

(1) promulgate a federal standard of performance for new sources pursuant to CAA Section

111(b)(1)(B); or (2) prescribe regulations pursuant to CAA Section 111(d)(1)(A) requiring states

to submit a plan “similar to Section 110” that adopts standards of performance for existing

sources. A source can be a “new source” or an “existing source,” but it cannot be both at the

same time. The Act’s binary division is then reemphasized in the definitions under Section

111(a), which provides:

42 U.S.C. § 7411(a)(2) – The term “new source” means any stationary source, the construction or modification of which is commenced after the publication of regulations (or, if earlier, proposed regulations) prescribing a standard of performance under this section which will be applicable to such source.

42 U.S.C. § 7411(a)(6) – The term “existing source” means any stationary source other than a new source.

The definitions are clear that an “existing source” cannot be a “new source.” Accordingly,

regulation of the same source under Section 111, subsections (b) and (d) at the same time cannot

be squared with the express language of the Act. “An agency has no power to ‘tailor’ legislation

to bureaucratic policy goals by rewriting unambiguous statutory terms,” regardless of how

11 Statement of House Energy and Commerce Committee Chairman Fred Upton at House Energy and Commerce Subcommittee on Energy and Power hearing (July 29, 2014). 12 79 Fed. Reg. at 34,876.



P a g e | 7

desirable the agency believes such an interpretation may be to supporting an agency’s proposed

rule.13

III. EPA LACKS THE LEGAL AUTHORITY TO DICTATE SUBSTANTIVE RESULTS UNDER SECTION 111(d), THIS IS THE PURVIEW OF THE STATES

With respect to the establishment of standards under Section 111(d), EPA’s authority is quite

circumscribed: EPA prescribes regulations establishing a state implementation plan-like

procedure under which the states submit to EPA a plan establishing standards of performance

for existing sources that reflect the degree of emission reductions each individual EGU can

achieve. Congress gave states the primary role in establishing standards of performance for

individual sources because states are best positioned to evaluate emission reduction potential in

the context of both economic and energy reliability concerns. The Agency does not have

authority under the CAA to dictate these substantive results. Though EPA claims that states

have flexibility to craft their own programs for compliance with the Proposed Rule, this

flexibility is illusory. In actuality, Arizona, among other states, for a multitude of reasons, would

face substantial challenges with achieving the rate-based carbon emission performance standard

proposed by EPA.

A secure and reliable grid and related infrastructure are important to protection of assets

safeguarded by the Department of Homeland Security. Ensuring there is a balanced portfolio

within each utility as well, as balanced locational assets, is important to ensuring a safe and

reliable grid, and it is the states that are best situated to uphold this vital responsibility. APS

urges EPA to withdraw this Proposed Rule and to propose a new rule that is specific to particular

sources, as is required by CAA Section 111(d), and to develop carbon emission standards that

utilities can achieve while still meeting their other federal requirements for electric grid

reliability and security.

13 UARG v. EPA, 573 U.S. ---, slip op. at 21 (2014).



P a g e | 8

IV. THE ARIZONA DEPARTMENT OF ENVIRONMENTAL QUALITY LACKS THE LEGAL AUTHORITY TO IMPLEMENT AND ENFORCE THE PROPOSED RULE

Arizona law currently prohibits an agency created by the governor, including the Arizona

Department of Environmental Quality (DEQ), from adopting or enforcing “a state or regional

program to regulate the emission of greenhouse gas [GHG] for the purposes of addressing

changes in atmospheric temperature without express legislative authorization.”14 The Arizona

DEQ may not be able to meet the proposed state plan submission deadline due to the time

needed for any requested legislative authorization before it can adopt a plan pursuant to EPA’s

Proposed Rule.

V. EPA FAILED TO PROPERLY CONSIDER A VARIETY OF POLICY ISSUES, INCLUDING EARLY-ACTION CREDIT, CONSTRUCTION TIMING, POPULATION AND LOAD GROWTH PROJECTIONS, AND INTERSTATE IMPLICATIONS OF ITS PROPOSED RULE

A. EPA’s State Goals Eliminate Any Purported State Flexibility, and EPA’s Methodology in Computing State Goals Is Flawed

EPA’s methodology in calculating each state’s, including Arizona’s, interim and final GHG

emission goals is flawed, and EPA must revise these goals. Furthermore, the state goals

eliminate any purported flexibility for states to meet their carbon reduction obligations.

The first major problem with EPA’s calculation methodology is that it applies its building blocks

approach to non-affected Subpart KKKK units. The proposed goals assume that states will

regulate these sources, but states can only regulate existing sources to which a New Source

Performance Standard (“NSPS”) would apply under Section 111(b).15 For GHG emissions, an

NSPS only applies to stationary combustion turbines that meet certain capacity-related criteria.16

14 Ariz. Rev. Stat. § 49-191. 15 42 U.S.C. § 7411(d)(1)(A)(ii). 16 Proposed 40 C.F.R. § 60.4305(c), 79 Fed. Reg. at 1,506. Specifically, only a newly constructed or modified stationary combustion turbine that on average supplies “one-third or more of its potential electric output capacity



P a g e | 9

Some natural gas combined cycle (NGCC) units to which generation would be shifted under

EPA’s Proposal for building block 2 do not qualify under these capacity-related criteria, and

EPA has not identified how many NGCC units in Arizona or elsewhere qualify under these

criteria. EPA’s calculations assumed all of these units, whether or not they are subject to

regulation, could participate in building block 2, arbitrarily and capriciously inflating the state

goals. Additionally, for those NGCC units that can properly be regulated, EPA made various

errors in calculating capacity, double-counting some units, and counting non-existing units.

These mistakes are explained in more detail by UARG’s consultant, James Marchetti, and APS

incorporates his report and UARG’s comments on this issue.17

The second major problem is EPA assumes that states can implement all Building Blocks, and

that if a state cannot fully implement one Building Block, it can increase another. However,

because most states cannot realistically implement at least one of the Building Blocks, for

instance because of the timing of installing new capacity, they have very limited ability to

achieve EPA’s goals. This is exacerbated by EPA’s faulty assumptions for many Building

Blocks (for instance, assuming full capacity of units that are still under construction), eliminating

any potential state flexibility in implementing the Building Blocks.

B. EPA Should Give Credit for Early Action by States that Have Already Implemented Greenhouse Gas Reduction Standards

Many states, including Arizona, have implemented mandatory energy efficiency (“EE”) and

renewable energy (“RE”) standards. EPA in the preamble to its Proposed Rule claims it does not

want to disadvantage states with existing programs, and APS agrees with that position.18

However, EPA’s Proposed Rule does, in fact, disadvantage states that have taken significant

action to reduce GHG emissions. EPA’s current Proposal is that only emission reductions

and more than 219,000 MWh as net-electric output to a utility power distribution” system is subject to the proposed CO2 NSPS. See id. 17 UARG Comments, Sections XIV.C.2 and XV.B. 18 79 Fed. Reg. at 34,918-19.



P a g e | 10

achieved during a state plan performance period as a result of actions taken after the Clean

Power Plan (“CPP”) was proposed can be used to demonstrate compliance with the state

performance target.19 However, EPA is taking comment on proposed alternatives, including

setting the date after which credit for early action can be obtained at a different time (as early as

2005), or allowing for credit for reductions achieved prior to the performance period. APS

encourages EPA to give the broadest favorable treatment for state actions taken to reduce carbon

emissions, and supports a 2005 baseline to facilitate this.20

Recognition of early action credit would best serve EPA’s stated policy goals of allowing

maximum flexibility and gradual carbon decreases. The use of a 2005 baseline would allow

states to build upon current programs and to use the demonstrated historical success of those

programs to comply with the interim goals to allow for a smoother transition. To maximally

facilitate existing plans and reductions under those plans, EPA should determine that any carbon

reductions achieved under state programs should be counted toward state compliance. For

instance, Arizona has successfully implemented both EE and RE measures that have had

quantifiable, surplus, enforceable, and permanent reductions in carbon emissions, and it should

receive credit for having achieved those reductions under EPA’s regulations.

C. EPA Failed to Consider How the Proposed Rule Interacts with Other Clean Air Act Rules and Affects Prior Investments in or Early Retirement of Fossil-Fired Units

In its Proposal, EPA failed to consider how its regulations will impact unit owners who have

modified or retired their fossil-fired units in response to state GHG requirements or federal CAA

rules. To the extent unit owners take early action to permanently retire higher-emitting fossil-

fired units (both prior to and after) finalization of the Section 111(d) regulations, they should be

allowed to take credit for any GHG reductions resulting from these closures, especially if they 19 79 Fed. Reg. at 34,918. 20 A 2005 baseline would be consistent with President Obama’s pledge that by 2020, the United States “would reduce its greenhouse gas emissions in the range of 17 percent below 2005 levels if all other major economies agreed to limit their emissions as well.” The President’s Climate Action Plan (June 2013), at 4.



P a g e | 11

took place after the proposed baseline year, 2012. EPA should include guidance in its Proposed

Rule to instruct states how they can include credit for these early reduction efforts in their

compliance plans.

D. EPA Did Not Adequately Account for the Time or Cost to Plan and Construct Transmission Lines or Gas Infrastructure That May Become Necessary

Aside from physical equipment costs, the transmission line siting process itself is costly and

time-consuming. Real estate negotiations (e.g., rights-of-way), environmental permitting

(including preparation of environmental assessments and environmental impact statements),

public opposition, and regulatory approval from various entities can take up to three years for

simple projects or up to ten years for complicated projects. Thus, it will generally be the case

that at least three years (and very potentially more) will have gone by before actual transmission

line construction can commence. EPA did not fully consider the impact of these significant

barriers, especially in states like Arizona and New Mexico, where large energy projects can

easily affect federal, state, tribal, and private lands simultaneously. Furthermore, these large

energy projects often have the potential to trigger a National Environmental Policy Act

environmental review and to impact threatened or endangered species or their critical habitat,

requiring Section 7 consultation under the Endangered Species Act, which delays these projects

even further.

E. EPA Ignored Population, Industrial Load Growth, and Peak Energy Demand Projections and Failed to Consider the Rule’s Interference with Economic Development

It is not at all clear whether EPA’s mandatory performance standards took into account future

potential load growth from population shifts (i.e., states losing or gaining population at

significant levels) or increases in the electrification of states’ transportation sectors.

Electrification of the transportation sector is widely viewed as an important means of reducing

GHG emissions from mobile sources, and EPA should not create a disincentive for states to

pursue these policies. However, load growth and peak energy demand increases from future



P a g e | 12

population shifts and increases in electrification of the transportation sector are not likely to be

entirely served by zero-carbon emitting electricity sources. Rather, increased energy demand

means that zero- and low-emitting sources will likely be operating at capacity, and states will be

forced to rely on higher-emitting sources to meet their increased generation needs. This could be

expected to increase the rate of carbon emitted by affected sources in the state above EPA’s-

mandated target. These factors significantly affect the energy demand but vary widely by state.

For example, Arizona is one of the fastest growing states in the nation, and will likely see

significantly increased energy demand by 2030. EPA’s failure to adequately account for these

factors could leave states unable to comply with their performance standards.

F. EPA Failed to Properly Consider the Interstate Effects of Its Proposed Rule

EPA’s proposed goals are state-based, yet (as the Agency acknowledges) the nation’s electricity

system operates across state and tribal lines. For instance, the Four Corners Power Plant near

Farmington, New Mexico is located on the Navajo Nation and is owned by participants

domiciled in Arizona (APS, Salt River Project, and Tucson Electric Power), New Mexico (Public

Service Company of New Mexico), and Texas (El Paso Electric). The Navajo Generating

Station, also located in Indian country, is owned by numerous private and public participants,

one of which is the U.S. Bureau of Reclamation. In fact, many utilities that will be affected by

the new Section 111(d) standards own and operate designated facilities located in multiple

jurisdictions, including states and tribal lands.

EPA’s Proposed Rule fails to provide clear ground rules on how these multi-jurisdictional

entities will be regulated and receive credit for reductions. While recognizing the complexities,

EPA in its Proposed Rule only proposes default rules on calculations for state plans as a whole,

but does not address how non-state entities that operate across state lines will be affected, nor



P a g e | 13

does EPA give guidance to states on how to address multi-jurisdictional entities other than

asking states to simply come to a mutual agreement.21

VI. THE PROPOSED RULE CONFLICTS WITH THE DIVISION OF FEDERAL AND STATE REGULATORY AUTHORITY PROVIDED UNDER THE FEDERAL POWER ACT

EPA’s interpretation of its authority to establish and enforce state energy programs in the

Proposed Rule conflicts with the division of regulatory authority between the federal and state

governments over electric utilities, pursuant to the Federal Power Act.22 The Proposed Rule

would impermissibly insert EPA into the separate areas of authority that the Federal Power Act

has assigned to the states and to the Federal Energy Regulatory Commission (“FERC”). EPA

should not adopt the approach set forth in the Proposed Rule because it is unlawful under the

Federal Power Act and would disrupt the current regulatory system. In addition, EPA’s

Proposed Rule is based on numerous factual errors and faulty assumptions relating to

dispatching, energy regulatory frameworks, energy infrastructure development, and transmission

constraints, in part due to its lack of coordination with FERC. These Agency errors run contrary

to the reasoned decision-making requirement under the Administrative Procedure Act.23

A. The Agency’s Proposed Rule Attempts to Usurp Control over Regulatory Matters that the Federal Power Act Leaves to the States

The preservation of state jurisdiction over generation, resource planning, and other matters under

the Federal Power Act is consistent with basic principles of federalism and the Tenth

Amendment. EPA’s Proposed Rule disregards the careful division of responsibilities provided

for in the Federal Power Act and attempts to claim for EPA portions of states’ authority.

Multiple aspects of the Proposed Rule would effectively result in the Agency making energy

21 79 Fed. Reg. at 34,921-22. 22 See 16 U.S.C. § 791a et seq. 23 5 U.S.C. § 706(2)(A).



P a g e | 14

policy decisions in areas that the Federal Power Act leaves to state control. Congress did not

empower EPA with this authority in enacting Section 111(d). In addition, EPA cannot have

greater authority over energy resources development and planning, which Congress expressly

reserved to state regulation, than FERC is permitted under the Federal Power Act.

B. The Proposed Rule Directly Conflicts with the Federal Energy Regulatory Commission’s Regulatory Authority

Sections 205 and 206 of the Federal Power Act empower FERC to regulate the interstate

transmission and wholesale sale of electric energy.24 However, the Proposed Rule would

interfere with these FERC-jurisdictional areas. For instance, the Agency’s proposed building

block 2—which provides for EGU carbon emissions reductions through re-dispatch from

affected steam EGUs to affected NGCC units25—conflicts with Federal Power Act in several

ways. EPA’s Proposed Rule effectively determines that the price increases needed to achieve a

70 percent NGCC utilization rate would be “reasonable,”26 but EPA lacks the authority to make

this determination, which comes within the purview of FERC’s authority. Building block 2

would also require changes to the dispatch and re-dispatch rules applicable to EGUs, but these

FERC-approved dispatch rules may only be amended as provided under the Federal Power Act.

EPA fails to acknowledge that modifying re-dispatch rules for environmental purposes, as

contemplated by Building Block 2, would require substantial changes to existing dispatch

procedures that could conflict with FERC policies. The Proposed Rule would require a

significant change to dispatching based on environmental considerations rather than the least-

cost economic dispatch rules. Even if the Agency’s proposed alteration to dispatch mechanisms

were technically feasible, which APS does not concede, it could only be approved by FERC, not

EPA.

24 16 U.S.C. §§ 824d, 824e. 25 79 Fed. Reg. at 34,856. 26 See 79 Fed. Reg. at 34,865.



P a g e | 15

Section 202(a) of the Federal Power Act authorizes FERC to divide the country into regional

districts for the voluntary interconnection and coordination of facilities for the generation,

transmission, and sale of electric energy, but such coordination must be voluntary and not

compelled by FERC. EPA’s Proposed Rule interferes with FERC’s jurisdiction by promoting

the development of multi-state plans for the coordinated operation and dispatch of EGUs and by

specifying rules to govern the development and evaluation of such plans. Though EPA claims

that any additional regional coordination would be voluntary, as a practical matter it appears that

the Proposed Rule effectively requires further coordination to reduce compliance costs and due

to the integrated nature of the interstate grid. EPA certainly cannot exercise greater power to

force regional coordination than the Federal Power Act allows FERC, nor can it indirectly render

mandatory regional coordination that Congress has otherwise determined must remain a purely

voluntary state option. The Agency’s Notice of Data Availability in support of the Proposed

Rule (“NODA”)27 further demonstrates EPA’s encroachment into the arena of energy regulation,

describing a regionalized approach to calculating renewable energy targets.28

The Proposed Rule also fails to adequately take into account the requirements of existing FERC-

approved reliability standards and how the obligation of transmission-owning utilities,

independent system operators and regional transmission organizations, and EGUs to comply with

them may conflict with the Agency’s proposed emission reduction goals. Federal Power Act

Section 215 provides FERC with the responsibility for reliability of the bulk power system

through its authorization and enforcement of reliability standards developed by the North

American Electric Reliability Corporation and regional reliability entities. EPA should work

with FERC and others with reliability expertise to modify the Proposed Rule as needed to

prevent conflicts between FERC-approved enforceable reliability standards and EPA

27 79 Fed. Reg. 64,543 (Oct. 30, 2014). 28 See 79 Fed. Reg. at 64,551. EPA’s NODA implicitly recognizes the fundamental misalignment between calculating building block 3 goals based on in-state generation and allowing compliance to incorporate renewable energy certificates.



P a g e | 16

requirements pursuant to the Proposed Rule. EPA should not be permitted to supplant FERC’s

reliability determinations with its own.

VII. EPA FAILED TO ADEQUATELY CONSIDER THE EFFECT OF NEW SOURCE REVIEW REQUIREMENTS ON IMPLEMENTATION OF THE PROPOSED RULE

Under the CAA’s New Source Review (“NSR”) rules, a major modification of a major source

triggers pre-construction permitting requirements. If the source is in a National Ambient Air

Quality Standards (“NAAQS”) attainment area, Prevention of Significant Deterioration (“PSD”)

rules require installation of Best Available Control Technology (“BACT”).29 If the source is in a

nonattainment area, then the Lowest Achievable Emissions Rate (“LAER”) must be met, and

emissions offsets must be obtained.30 “Modification” includes any physical change or change in

method of operation of a stationary source which increases emissions of any air pollutant.31

Implementation of the Proposed Rule’s building block 1 on-site efficiency measures is very

likely to trigger stringent NSR requirements.32 Increasing a plant’s efficiency may also increase

its utilization, which may lead to a net increase of one or more criteria pollutants, thereby

triggering NSR. A source would then be required to go through the NSR permitting process and

install BACT (or LAER and offsets, if applicable) in addition to the envisioned efficiency

upgrades to comply with Section 111(d). The prospect of a burdensome regulatory process

could chill the very efficiency improvements the Section 111(d) program is designed to

encourage.

EPA’s Proposal does not resolve the conflict between the Proposed Rule and NSR. Rather, EPA

states that it does not expect many instances where “an NSR permit would be required” when

29 See 42 U.S.C. §§ 7470-7479. 30 See id. 7502(c)(5), 7503. 31 Id. §§ 7479(2)(C), 7411(a)(4). 32 §§ 7479(2)(C), 7411(a)(4).



P a g e | 17

implementing the Rule’s requirements.33 However, a review of responses to efficiency

improvements over the past decades shows that improvements of the kind suggested by EPA

have frequently triggered either regulatory enforcement or citizen litigation under NSR

standards.34 As a result, the Proposed Rule likely will lead to inconsistent or duplicative

regulation that imposes additional costs and procedures on regulated entities. The President has

directed federal agencies, including EPA, in Executive Order 13,563 (Jan. 18, 2011) to

harmonize regulations to avoid “redundant, inconsistent, or overlapping” regulatory

requirements. Consistent with this order, EPA should harmonize its Section 111(d) Proposal

with NSR in its final rule.

APS suggests that EPA develop appropriate alternative screening methods for an expedited,

streamlined NSR process. Such methods can be designed to ensure that any criteria pollutant

increase would not interfere with the attainment of NAAQS. These screening tools can easily

confirm that well-controlled sources and sources with net emissions increases of criteria

pollutants below attainment-related significance thresholds (i.e., levels that would interfere with

attainment) have satisfied NSR requirements. EPA could also find that compliance with Section

111(d) standards satisfies applicable PSD and BACT requirements for GHG emission reductions,

since the standards should represent the maximum degree of reduction achievable, taking into

account energy, environmental, and economic impacts and other costs.35

VIII. EPA RELIES ON AN OVERESTIMATE OF PURPORTED BENEFITS TO JUSTIFY THE PROPOSED RULE

The Agency has overestimated the benefits of the Proposed Rule in both its monetization of the

“social cost of carbon” (“SCC”) and public health benefits. Since the CAA requires that the cost

of achieving the reductions through the BSER be taken into account in establishing standards of

33 79 Fed. Reg. at 34,859. 34 UARG Comments, Section XII. 35 See 42 U.S.C. § 7479(3) (definition of BACT).



P a g e | 18

performance,36 a more accurate estimate of the Proposed Rule’s purported benefits is needed to

ensure future actions taken on the Proposal are justified under the statute and as a matter of

sound public policy.

A. Social Cost of Carbon

APS objects to EPA’s use of SCC estimates to justify this regulation, and incorporates the

comments submitted by UARG on EPA’s Proposed Rule in UARG Comments37 on the Office of

Management and Budget’s (“OMB”) request for comments on the U.S. government’s

development of SCC values.38 The SCC is an estimate of the monetized global climate change

damages associated with an incremental increase in carbon emissions in a given year. The SCC

protocol was developed by an interagency working group as a cost-benefit analysis tool in

response to Executive Order 12,866, which requires agencies to assess both the anticipated

benefits and anticipated costs of a proposed economically significant regulatory action.

There are multiple problems with EPA’s use of the SCC estimates to justify the Proposed Rule.

First, there is no consensus on a single discount rate for the SCC estimates, which is why the

interagency working group suggested four different approaches in the SCC 2010 Technical

Support Document,39 resulting in a very wide range of cost estimates. The interagency working

group decided against using the 7 percent discount rate, which is typically used for agency rule-

makings.40 This rate would present a lower, more reasonable social cost of carbon estimate.

36 42 U.S.C. § 7411(a)(1). 37 UARG Comments, Section XIII. 38 UARG comments to OMB on the Notice of Availability and Request for Comments on the Social Cost of Carbon Technical Support Document and Updates, 78 Fed. Reg. 70,586 (Nov. 26, 2013), Docket ID No. OMB-2013-0007-0091 (Feb. 26, 2014). 39 See Interagency Working Group on Social Cost of Carbon, “Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis under Executive Order 12866” 23 (Feb. 2010). 40 The OMB guidance specifies that agencies should use both a 3percent and a 7percent discount rate when determining costs and benefits of rule-makings, although acknowledging that lower discount rates might be appropriate where there are intergenerational costs or benefits. OMB, Circular A-4 (Sept. 17, 2003), at 34. In the



P a g e | 19

Second, the SCC protocol estimate is not meaningful, in part, because it does not take into

account the social cost of a replacement source of electricity needed to meet demand if a source

is shut down. The GHG emissions of replacement electricity are unknown, but are not zero and

could even be higher depending on the replacement source. Third, since the SCC estimates

incorporate the global costs of domestic carbon emissions, EPA’s use of the SCC estimates

shows international benefits—not just domestic benefits—from domestic carbon emission

reductions. It is not appropriate for EPA to rely on global benefits based on the SCC estimates

because EPA is weighing these global benefits against domestic costs, and the Agency lacks

authority to control the potential climate benefits that would take place outside the United States.

B. Public Health Benefits

EPA improperly attributes public health benefits to the Proposed Rule in two main ways. First,

the Agency credits the Proposed Rule with public health benefits from reductions in fine

particulate matter (“PM2.5”) and ozone (“O3”), which are already mandated by existing

regulatory programs, including the NAAQS program. EPA already set the NAAQS at levels that

the Agency determined are protective of public health, and which incorporate an adequate

margin of safety. The Agency cannot fairly credit the Proposed Rule with the health benefits of

emissions reductions that will occur regardless of the implementation of EPA’s Section 111(d)

Proposal. Second, EPA attributes public health benefits to the Proposed Rule that stem from

reductions in concentrations of PM2.5 and O3 to levels below what is already required under the

NAAQS. It is improper for the Agency to credit health benefits to the Proposed Rule based on

exposure to PM2.5 and O3 at levels below the thresholds that EPA has already deemed protective

of public health.

Overall, these flaws in the Agency’s calculations of the SCC and public health benefits

significantly overestimate the asserted benefits of the Proposed Rule.

Proposed Rule, EPA calculated other non-GHG-related costs and benefits using the 3percent and 7percent discount factors. See 79 Fed. Reg. at 34,839.



P a g e | 20

IX. ARIZONA’S INTERIM AND FINAL GOALS ARE AMONG THE MOST STRINGENT IN THE NATION AND ARE NOT REPRESENTATIVE OF TYPICAL OPERATIONS

A. Arizona is Penalized for its Forward-Thinking Generation Mix with One of the Most Stringent Goals in the Country

Notwithstanding the interrelationship that exist between EPA’s building blocks, EPA

independently applied each building block to Arizona’s 2012 baseline carbon emission rate to

systematically reduce the carbon emission rate and propose an interim goal of 735 lb/MWh

during the 2020 through 2029 time frame and a final of 702 lb/MWh by 2030. The interim goal

represents a 47 percent reduction and the final goal represents a 52 percent reduction from the

2012 level. While Arizona is expected to achieve a 52 percent reduction from the baseline

emission rate, 80 percent of the states are required to achieve less than a 40 percent reduction.

When Arizona’s emission reductions are compared to EPA’s purported 30 percent reduction

from the power sector’s 2005 carbon emission level, it is unclear why a small percentage of the

nation is proposed to carry the heavy burden of reducing the nation’s carbon emission rates.41

Arizona must achieve disproportionately larger emission reduction goals than a vast majority of

other states across the country and a much larger reductions than all the states neighboring

Arizona mostly because the Proposed Rule assumes Arizona has a relatively large amount of

“underutilized” existing NGCC capacity. Based on that assumption, EPA suggests that Arizona

should simply displace all coal-fired generation irrespective of technical limitations and

economic dispatch concerns. In contrast, if a state had not already made efforts to diversify its

generation portfolio to take advantage of zero or lower carbon-emitting generation sources such

as renewable and NGCC, then that state’s reduction requirement to lower the nation’s carbon

footprint would be inexplicably small. APS believes that EPA’s approach to setting the states’

carbon emission rate goals penalizes Arizona for a progressive, diverse generation mix.42

41 79 Fed. Reg. 34,832, 34,839 42 Section XIII of these comments provides additional details regarding the economics of the stringent goals.



P a g e | 21

B. APS Does Not Support the Use of 2012 as the Baseline Year

Carbon emissions from an NGCC unit are approximately 60 percent lower than those from a

coal-fired EGU.43 Thus, it is not a surprise that EPA is seeking to shift generation from coal-

fired EGUs to more natural gas-fired EGUs. However, this carbon emissions savings rate

through re-dispatch is a fundamental reason why EPA should not use 2012 to calculate baseline

emissions rates. A key benchmark location for natural gas pricing throughout the United States

is at Henry Hub (HH) in Erath, Louisiana. Reviewing HH natural gas (pipeline) pricing data

from 2002 to 2012, the electric utility industry benefitted from the lowest natural gas prices in a

decade. In fact, in a single year, from 2011 to 2012, natural gas prices dropped 32 percent.44

This anomalous low price point was predicated on the fact that demand was far behind a steadily

increasing production of shale gas resulting in an abundant natural gas supply in 2012.45

Because EGUs operate on an economic dispatch model, and not an environmental dispatch

model, generation shifted from coal-fired EGUs to natural gas-fired EGUs during 2012. From

EPA’s formula to calculate the 2012 “unadjusted fossil emission rate (or historical fossil rate),” it

is obvious to see how a significant, albeit temporary, shift in generation from coal-fired to

natural gas-fired would result in a baseline carbon emission rate that is much lower than would

be calculated under normal source operations.46

EPA’s use of the term “unadjusted fossil” highlights the disparities between the selection of a

more representative year versus the selection of a year (2012) under which the nation’s utilities

experienced atypical operations. Although any EGU may experience irregularities based on

location factors (e.g., weather) or situational factors (e.g., forced or planned outages), it is not

often that an entire sector, such as the energy sector, is affected by a single factor like the

43 Jackson Salovaara, Harvard Environmental Economics Program, Coal to Natural Gas Fuel Switching and CO2 Emissions Reductions, Discussion Paper 11-27 (2011). 44 www.eia.gov/dnav/ng/hist/n9102us3a.htm. 45 American Gas Association, U.S. Natural Gas Supply: Understanding Abundance in 2012 – Just the Facts, Energy Analysis, EA 2011-09 (Dec. 14, 2011). 46 EPA, Goal Computation Technical Support Document, Docket ID No. EPAHQ-OAD-2013-0602 (June 2014).



P a g e | 22

astonishingly low (and unprecedented) natural gas prices. Conversely, as natural gas prices have

subsequently risen, utilities have increased the dispatch of coal-fired units.

EPA states the following in the preamble of the Proposal: “Nationwide, by 2030, this rule would

achieve carbon emission reductions from the power sector of approximately 30 percent from

carbon emission levels in 2005.”47 This is important to note because EPA selected 2012 as the

baseline year in the Proposed Rule. Not only is this particular year not representative of typical

operations during this period in history, as discussed above, the selection of 2012 also negates,

for the purposes of calculating carbon emission reductions, the progressive moves made by the

power sector through heat rate improvements, re-dispatch to NGCC units on a more widespread

basis (not just via economic dispatch over a single year), and the development of robust RE and

EE programs initiated by the electric utility industry and assisted by the Energy Policy Act of

2005.48 If these activities look familiar, they should. They are the same building blocks that

EPA used in the Proposed Rule. APS believes that electric utilities should be able to take credit

for carbon emission-reducing activities by the use of a baseline year that is more representative

of typical operations and which goes back to at least 2005.

In EPA’s Proposed Rule, Table 3 (U.S. GHG Emissions and Sinks by Sector) demonstrates the

impact that a nationwide reliance on more NGCC versus coal-fired generation can potentially

have on carbon emissions.49 Carbon emissions from the energy sector alone were 12 percent

higher in 2005 than in 2012. Notably, overall carbon emissions from this same sector increased

only 4 percent from 1990 to 2012. EPA’s own data, combined with the information regarding

how fossil fuel-fired EGUs were dispatched on a nationwide basis in 2012 demonstrates that the

carbon emissions in 2012 were artificially low and not representative. Accordingly, it is

inappropriate to use 2012 as a baseline for carbon emission rate reductions.

47 79 Fed. Reg. at 34,832. 48 Public Law, 109-58, 109th Congress, Energy Policy Act of 2005, Aug. 8, 2005. 49 79 Fed. Reg. at 34,843.



P a g e | 23

X. GENERAL GOAL SETTING ISSUES

In the Proposed Rule, EPA applies its newly developed BSER for existing fossil fuel EGUs to

determine state-specific, rate-based goals for carbon emissions from the power sector. In order

to establish and apply BSER, EPA made numerous assumptions. APS closely reviewed EPA’s

assumptions and associated analyses and identified many technical errors and false assumptions

used by EPA in establishing Arizona’s carbon emission rate goals. As a result of these technical

errors and incorrect assumptions, EPA proposed carbon goals for Arizona that are unachievable,

and any attempt to achieve these goals would create significant reliability concerns for Arizona’s

electric energy supply and result in a substantial cost increase to Arizona customers.

A. Compliance Flexibility is a Misconception

EPA explains in the Proposed Rule that each state will have “flexibility to design a [111(d)]

program to meet its goal in a manner that reflects its particular circumstances and energy and

environmental policy objectives.”50 EPA also states that it established each building block using

conservative estimates to allow states that are forced to underachieve in one building block to be

able to correct the deficit by overachieving in another building block.51 In practice, however,

this is a huge misconception; EPA’s promise of “flexibility” rings hollow. For example, EPA

used the re-dispatch of all of Arizona’s coal-fired EGUs to NGCC units to establish the proposed

Arizona goals. When EPA was asked how Arizona could possibly accomplish such a feat

without creating a severe challenge to electric reliability in the state, EPA responded by saying

that Arizona may use other means of achieving the state goals in lieu of the re-dispatch of coal-

fired generation to NGCC units. For example, the state may employ more RE and EE

requirements, EPA explained.

However, the Arizona Department of Environmental Quality (ADEQ) analyzed increasing RE

and EE standards in the state to allow for the continued operation of a portion of the coal-fired

50 Id. at 34,833. 51 Id. at 34,926.



P a g e | 24

fleet in Arizona and, in fact, these actions have the opposite effect.52 In the ADEQ comments to

EPA, the State postulated various scenarios of increasing the incremental contribution from RE

and EE so as to reduce the amount of re-dispatch from coal to NGCC in order to preserve some

coal-fired generation in the state. When ADEQ performed this analysis they discovered “the

more Arizona relies on BB3 and BB4 to achieve compliance with the final goal of 702 lbs

CO2/MWh, the farther it gets from compliance with the interim goal (scenarios 2-4). Conversely,

if Arizona designs its program to comply with the interim goal, it ends up with a final rate that is

far lower than necessary to comply with the final standard and ends up preserving a much

smaller portion of its existing coal-fired generation resources.”53 Therefore, the State is left

without an effective way to use flexibility between the building blocks to achieve the proposed

Arizona goals.

EPA also suggested that Arizona’s coal-fired units could be operated as peaking plants to cover

the peak energy demand during the summer months. While such an option may be available at

facilities that have a variety of EGUs that are co-located and can combust different fuels, the

majority of the generating stations with coal-fired EGUs in Arizona only have coal-fired units.

Therefore, operating the Arizona coal-fired EGUs only during the summer months is unrealistic

from an economic perspective. It is not economically justified to keep a full operating

complement on site for only several months of plant operation during the year.

B. Feasibility Challenges for Achieving Proposed Interim Goals

Arizona faces serious feasibility problems with achieving the interim goal proposed by EPA,

particularly due to the front-loaded nature of EPA’s proposed interim goal for Arizona. To

comply with the state’s interim goal, more than 90 percent of Arizona’s total reductions in

carbon emissions must occur by 2020. According to Arizona DEQ’s calculations, if Arizona

designs its program to comply with EPA’s proposed interim goal the state would overshoot its 52 Arizona Department of Environmental Quality Comments on Building Block 2 submitted to Docket ID No. EPA-HQ-OAR-2013-0602 (Nov. 21, 2014). 53 Id.



P a g e | 25

final goal by constructing more renewable generation and relying on increased EE, ending up

with a final rate that is much lower than needed to comply with the proposed final goal.54

Therefore, switching from coal to natural gas by 2020 is the only building block available to

Arizona to meet EPA’s proposed goal. Correspondingly, re-dispatching all coal renders building

block 1 meaningless for Arizona. Because of all the technical issues associated with the re-

dispatch of coal-fired generation to NGCC units by 2020, Arizona is faced with insurmountable

challenges in meeting the proposed interim goals. This is contrary to EPA’s stated intent that

states be afforded flexibility among the building blocks to achieve the proposed goals.55

C. Emissions Rates for Existing NGCCs Should Not Be Higher Than for New NGCCs

The existing NGCC units should not be held to a performance standard more stringent that the

performance standard for new NGCC units. EPA used a single year to depict the carbon

emissions rates for existing NGCC units. Apparently, EPA assumed all future years will be

consistent with the single, baseline year. This assumption is not necessarily true for at least a

couple of reasons. First, equipment wears over time and efficiencies are lost over the life of the

equipment. EPA recognized this phenomenon when it proposed the carbon emission rate

standards for new NGCC units. In that proposal, EPA states “These values represent the

emission rates that modern high efficient NGCC facilities located in the U.S. would be able to

maintain over its life (emphasis added).”56 Second, the single baseline year is not necessarily

representative of future operations. As more renewable generation comes on line, existing

NGCC units will be doing more load following. When operating in such a continuous load

following manner carbon emission rates will creep higher. Therefore, when establishing the

baseline for carbon emission rate goals, EPA should not use a rate lower than what is expected

for new, efficient NGCC units.

54 See Comments of Steve Burr, Arizona Department of Environmental Quality (Aug. 22, 2014). 55 Id. 56 79 Fed. Reg. 1487.



P a g e | 26

D. Interim Goals Are Unnecessary

EPA proposed its interim carbon emission rate goals as a means of assuring reasonable progress

towards achieving the final carbon emission rate goal. This is the only real purpose of the

interim goals, but it is unnecessary. As pointed out by EPA, utilities have been, and will

continue to take actions every year to reduce carbon emissions.57 Moreover, with a looming

final carbon emissions goal, the utilities will be motivated to continue work towards achieving

the final goal. Because the final carbon emission rate goal will be incorporated into the

operating permits of the affected utilities, there will be sufficient motivation to assure

compliance with the final goals by 2030. Furthermore, the difference between the carbon

emission reduction that would occur without interim goals and the carbon emission reduction

that will occur with interim goals will do little to affect climate change, rendering the interim

goals unnecessary. Because of these reasons APS does not support the use of interim goals for

achieving the desired outcome of the Proposed Rule.

However, if EPA determines that interim goals must be used, at the very least EPA should allow

the states to set the interim goals in order to establish a logical compliance trajectory for each

individual state. Each state will have a better understanding of what actions will be necessary to

achieve the final goals and the appropriate timing for achievement. For example, APS is

currently pursuing actions with the state that will drastically reduce future carbon emissions, but

due to certain requirements, these actions cannot be completed until the mid-2020s. This is

exactly the type of information the states will have to use in setting more realistic and achievable

interim targets.

The currently proposed interim goals for Arizona are too heavily weighted toward the early

years. EPA presumes that these goals can be achieved by re-dispatching all coal-fired generation

in the state to NGCC units commencing in 2020. As further detailed in the discussions regarding

re-dispatch in Arizona, this change is not achievable as envisioned by EPA. There are

57 79 Fed. Reg. at 34,830, 34,847, 34,904.



P a g e | 27

significant concerns regarding electrical reliability and infrastructure that prohibit the total re-

dispatch of the coal-fired fleet to NGCC units by such an aggressive time. If EPA is unwilling to

allow the states to set the interim goals, EPA must be mindful of the fact that Arizona’s utilities

need a more level trajectory that provides additional time to fully prepare and respond in a

reasonable fashion to the actions necessary to comply with any final goal. Accordingly, APS

supports changes to the Proposed Rule that make the 2020 to 2029 “glide path” more gradual,

such as a phase-in of building block 2 as EPA presents in the October 2014 NODA.58 EPA

should take into account whether, and to what extent, additional infrastructure improvements,

including natural gas pipeline expansion and transmission improvements, are needed when

developing the goals for building block 2. Such an approach would be consistent with the North

American Electric Reliability Corporation’s (“NERC”) conclusion that “more time would be

needed in certain areas to ensure resource adequacy, reliability requirements, and infrastructure

needs are maintained.”59

The wide variation in the difference between the proposed interim and final goals on a national

basis is arbitrary and capricious. As long as a state has prepared a 111(d) plan documenting

compliance with a reasonable final goal, interim goals are unnecessary.

E. Final Goal for Arizona is Inappropriate

Arizona will face a difficult dilemma as a result of the flawed assumptions used by EPA when

proposing the Arizona goals. The re-dispatch of all of Arizona’s coal-fired generation to NGCC

units is fraught with insurmountable technical difficulties, especially within the interim period.

Even if such an outcome could eventually be achieved, the result leaves the state’s electric

energy supply system vulnerable due to a diminished fuel mix and will substantially increase

energy costs to Arizona customers due to having to construct the necessary supporting 58 79 Fed. Reg. at 64,548-49. 59 North American Electric Reliability Corporation, Potential Reliability Impacts of EPA’s Proposed Clean Power Plan: Initial Reliability Review, at 27 (Nov. 2014) [hereinafter “NERC Report”] (“The EPA, FERC, the DOE, and state utility regulators should consider their regulatory authority to make timing adjustments and to grant extensions to preserve [bulk power system] reliability.”).



P a g e | 28

infrastructure. The most likely outcome is that the state’s utilities will be left with the difficult

decision of whether to jeopardize electric reliability in Arizona or risk noncompliance with the

proposed carbon emission rate reduction goals. Clearly, creating an unreliable electrical system

and substantially increasing energy costs in Arizona is not EPA’s intention or desired outcome;

therefore, it is imperative that Arizona’s final goal be realistic and not create an unreliable energy

supply system or lead to unaffordable energy.

F. Alternative Goals

EPA proposed an alternate set of emissions goals and has asked for comments on the

appropriateness of the alternative goals.60 APS does not support the adoption of any of the

alternative interim or final goals for Arizona. Most of the problems associated with achieving

the Arizona goals proposed by EPA surround the short duration for their achievement. An even

shorter time period to achieve the goals, albeit somewhat higher, would only exacerbate these

problems.

XI. PROPOSED EFFICIENCY IMPROVEMENTS FOR SUBPART Da EGUS ARE UNACHIEVABLE

APS supports the use of heat rate efficiency improvement projects at existing coal-fired EUGs

(Subpart Da units) as a means of reducing the carbon emission rate. Heat rate efficiency

improvement projects will achieve lower carbon emission rates and may be cost justified. In

fact, such actions are business as usual for electric utilities, as utilities have been implementing

efficiency improvement actions for decades. As a legal matter, building block 1 is the only legal

method pursuant to Section 111(d) for reducing carbon emissions at existing Subpart Da units.

However, the performance standard must be achievable at all units based on measures that can be

integrated into the design and operation of the source. Contrary to EPA’s assertion, based on the

professional judgment of APS, which is supported by years of engineering and operational

experience with Subpart Da units and corroborated by current research, a six percent efficiency

60 79 Fed. Reg. at 34,839.



P a g e | 29

improvement for all Subpart Da units is simply unachievable. Therefore, APS does not support

the required six percent heat rate improvement on Subpart Da EGUs as proposed by EPA.

EPA relies on a combination of best operating practices (four percent) plus equipment upgrades

(two percent) to collectively establish the six percent efficiency improvement target for each

affected Subpart Da EGU.61 The carbon emission rate reduction associated with the efficiency

improvements at each Subpart Da EGU was then carried forward and used to set the interim and

final carbon emission rate goals. The six percent improvement proposed by EPA is unfounded

and states cannot reliably count on this magnitude of efficiency improvement on every Subpart

Da source to comply with the state goals.

A. Efficiency Improvements from Operating Practices

The methodology used by EPA assumed the variability in heat rate of EGUs is directly

correlated to operating practices of the units.62 However, EPA doesn’t identify a specific

efficiency gain that corresponds to a specific operating practice. EPA does point out operating

practices, such as, shutting down unneeded pumps, installing digital control systems, more

frequent tuning of existing control systems, and more frequent change out of like-kind

replacement components, but EPA does not identify a corresponding efficiency gain by these

actions.63 In the Proposed Rule, EPA states:

Assuming that between 10 percent and 50 percent of the deviation from top decile performance in each subset of hourly heat rate observations within defined ranges of temperature and load could be eliminated through adoption of best practices, the result is a corresponding estimated range of 1.3 percent to 6.7 percent technical potential for improvement in the average heat rate of the entire fleet of coal-fired EGUs.64

61 79 Fed. Reg. at 34,860. 62 Id. 63 Id. 64 Id. (emphasis added).



P a g e | 30

Based on EPA’s language, it is clear that EPA is uncertain whether a four percent efficiency gain

can be achieved with better practices. APS is certain, however, that it cannot achieve this

magnitude of efficiency gain on the EGUs operated by APS because APS already employs the

industry’s best operating and maintenance practices on our coal-fired units. Therefore, there is

little to no room for additional efficiency gains in this area. This means in order to achieve an

overall six percent efficiency improvement, more efficiency gains would have to be achieved

through equipment upgrades, in which there is no certainty that even the two percent gain

proposed by EPA can be achieved.

B. Efficiency Improvements from Equipment Upgrades

EPA relies on a 2009 Sargent & Lundy study65 to show a proposed two percent efficiency

improvement for equipment upgrades is achievable. Based on a literature-survey-and-vendor-

data analysis, EPA uses national averages to conclude that existing coal-fired EGUs can achieve

“an aggregate heat rate improvement of approximately four percent (incremental to the

improvement achievable for adoption of best practices).”66 EPA goes on to explain that it

recognizes that some of the efficiency projects identified in the Sargent and Lundy study have

already been implemented and, therefore, concluded that a two percent efficiency improvement

due to equipment upgrades at Subpart Da EGUs is reasonable. However, as previously stated,

APS will not be able to achieve a four percent efficiency improvement based solely on operating

and maintenance practices, and therefore the overall six percent efficiency gains must be

achieved through equipment upgrades. Opportunities to achieve this level of efficiency gain

based on equipment upgrades at APS simply do not exist.

1. Projects Have Already Been Done

Sargent and Lundy based its results on a survey of power plants to determine heat rate

improvement projects that had already been completed. While the study could be used to identify 65 Coal-fired Power Plant Heat Rate Reductions, SL–009597 Final Report, January 2009, available in the rule-making docket and at http://www.EPA.gov/airmarkets/resource/docs/coalfired.pdf. 66 79 Fed. Reg. at 34,860.



P a g e | 31

potential heat rate improvement projects, it also can be used to demonstrate that many of the heat

rate improvement projects have already been performed. The data used in the Sargent and Lundy

study are over six years old. Many of the suggested upgrades had already been performed prior

to the 2012, which is the baseline year used in the building block 1 calculation. EPA recognizes

many of the projects have already been conducted, as EPA provides this as justification for only

requiring two percent efficiency improvement from equipment upgrades.67 This recognition by

EPA demonstrates that national averages cannot be used to set a fixed efficiency improvement

target. Applying national averages directly to individual units does not work. Each unit must be

evaluated on a case-by-case basis to determine what additional efficiency improvements can be

made to the unit. A specific unit’s heat rate is primarily due to factors outside of the source’s

control and is tied to such factors as plant location, design, age, and size, coal type, cooling

system type, grid dispatch, time of year the unit is dispatched, ambient temperature, auxiliary

load requirements, pollution controls, and other factors that EPA has not holistically considered.

EPA did not conduct a case-by-case analysis to determine what future projects are possible for

all coal-fired power plants. In fact, the two specific plant types Sargent and Lundy studied in

order to quantify its heat rate improvements—a 1970s vintage 250 MW boiler with an

electrostatic precipitator and a late-1970s vintage 850 MW boiler with a baghouse and

scrubber—do not currently exist in the U.S. operating coal-fired plant fleet.68

2. Illusionary Efficiency Gains

In EPA’s GHG Abatement Measures Technical Support Document, EPA points to large heat rate

improvements via equipment upgrades that could be replicated on other EGUs.69 However, in

many of the cases where EPA believed an efficiency improvement occurred, the reality is the

67 Id. 68 Sargent & Lundy, Coal-Fired Power Plan Heat Rate Reductions, SL-009597, Final Report (Jan. 22, 2009). 69 Technical Support Document (TSD) for Carbon Pollution Guidelines for Existing Power Plants: Emission Guidelines for Greenhouse Gas Emissions from Existing Stationary Sources: Electric Utility Generating Units; GHG Abatement Measures (June 2014).



P a g e | 32

perceived efficiency improvement was actually just the result of a change in continuous emission

monitoring system reporting methodology and not the result of an equipment or operational

change to lower heat rate.70 The actual heat rates did not improve. Rather, a more accurate

assessment of air flow through the stack caused the units to report an apparent improved heat

rate.

3. Projects May Only Provide Short-Term Benefits

EPA has not considered that the projects cited in the Sargent & Lundy study do not account for

the fact that equipment wears over time resulting in degraded efficiency improvements. Any

estimate of a heat rate improvement is only for a given point in time, and the effectiveness of the

equipment upgrades tends to degrade over time as the equipment wears. Many of the efficiency

improvement projects in the Sargent & Lundy study result in diminishing efficiencies. For

example, the best year may not be representative of the entire unit’s outage cycle. The unit may

operate most efficiently the year after the outage and degrade until the next outage.

4. Impact of Parasitic Loads

The operation of sulfur dioxide scrubbers, selective catalytic reduction (SCR) controls, and

fabric filter degrades the heat rate of EGUs. To reduce emissions, utilities have been installing

and will continue to install such pollution control equipment. Between 2002 and 2012, a total of

120,000 MW of capacity was fitted with flue gas desulfurization (FGD) and 140,000 MW of

capacity was fitted with SCR.71 These projects result in parasitic loads that have a negative

effect on unit efficiency, offsetting the efficiency gains due to heat rate improvement projects.

For example, an FGD retrofit may decrease overall boiler efficiency by 1 to 2 percent, whereas

an SCR retrofit may decrease boiler efficiencies by up to 1.5 percent.72 The efficiency loss from

70 Critique of EPA’s Use of Reference Units to Select Heat Rate Reduction Targets, J. Cichanowicz and M. Hein (Oct. 2014). 71 Evaluation of Heat Rate Improving Techniques for Coal-Fired Utility Boilers as a Response to Section 111(d) Mandates, J. Cichanowicz and M. Hein (Oct. 2014). 72 Id.



P a g e | 33

such projects is cumulative; therefore, a unit retrofitted with both an FGD and an SCR could lose

up to 3.5 percent efficiency.

APS is required to install pollution control equipment on coal-fired boilers in the near future to

meet Mercury and Air Toxics Standards and regional haze requirements. Heat rate degradation

due to future compliance requirements must be accounted for prior to any heat rate improvement

requirement. Accordingly, the six percent heat rate improvement that EPA used to set the goals

is arbitrary and unrealistic. The heat rate value used to set the goals should be based on real net

improvements. Each coal plant should be assessed considering the actual heat rate improvement

projects that can be completed, discounted by the projects required by other environmental

regulations that will create parasitic loads. The aggregate of the real net heat rate improvement

projects, discounted for degradation, should then be used to set the state goals.

Of the efficiency improvement projects suggested by EPA for coal fleet implementation, many

of the projects would trigger NSR and likely could require the implementation of BACT

requirements.73 Adding pollution control devices to meet BACT requirements is another source

for parasitic loads. So most, if not all, of the efficiency gained from the equipment upgrade

project could be lost due to the additional pollution control equipment that would be needed to

implement the project.

5. Efficiency Gains are Lost at Reduced Loads

Another detrimental impact to the coal-fired EGUs’ heat rate is increased cyclic and part-load

operation. Many boilers today already operate in a load-following mode for at least a portion of

their operation. APS has experience with this recent phenomenon with our coal-fired fleet. With

an increase of non-dispatchable renewable energy connected to the electric grid, coupled with the

re-dispatch scenario required under building block 2, coal-fired units have been, and will likely

73 Technical Support Document (TSD) for Carbon Pollution Guidelines for Existing Power Plants: Emission Guidelines for Greenhouse Gas Emissions from Existing Stationary Sources: Electric Utility Generating Units; GHG Abatement Measures; June 2014



P a g e | 34

continue to be, operated at reduced loads. Cyclic and part-load operation of coal-fired plants has

a negative effect on unit efficiency, resulting in yet another offset to any potential efficiency

gains to heat rate improvement projects. A typical pulverized coal plant will experience an

increase in net plant heat rate of 6-10 percent when the load is decreased to 50-60 percent of

maximum plant output. Figure 1 below shows the typical impact to efficiency when boilers are

operated at reduced loads.

Figure 1

Heat Rate Impacts versus Part-Load Operation for a Coal-Fired Power Plant



P a g e | 35

C. Efficiency Improvements APS Potentially Could Achieve

APS operates a fleet of six coal-fired EGUs at two separate electrical generating stations. The

Four Corners Power Plant has two 770 MW (nominal) supercritical boilers that are equipped

with FGD and fabric filters, and soon will be equipped SCR. The Cholla Power Plant has four

tangentially fired boilers ranging in size from 116 MW to 380 mw (nominal). Three of the four

Cholla units are equipped with FGD and fabric filters. APS conducted a study to determine the

heat rate improvement projects that could be implemented on these units to effect efficiency

improvement and identified a few projects that could be implemented to achieve an improved

heat rate. However, when combined with other required environmental projects that will

negatively impact unit heat rate, such as the installation of SCR controls and a stack liner, the net

effect is a potential maximum efficiency improvement of between 0.7 percent and 1.6 percent.

The efficiency improvement study at Cholla showed that an efficiency improvement between 0.4

percent and 2.1 percent may be achievable. While APS agrees that these projects will be helpful

in lowering the carbon emission rate, the proposed six percent efficiency improvements are

simply not available for these units.

In summary, the contribution to the state’s carbon goal from efficiency actions (operating

practices and equipment upgrades) at Subpart Da EGUs should be determined by each state after

a careful case-by-case analysis of what can actually be achieved at each EGU. Not only will this

result in more realistic and achievable carbon emission rate reductions, this is precisely what the

CAA requires.

XII. RE-DISPATCH AS ENVISIONED BY EPA CANNOT BE ACHIEVED

Under building block 2, EPA uses changes to dispatch among the affected EGUs to establish the

state-specific emission rate interim and final goals. Based on building block 2, EPA proposes to

reduce carbon emission rates by shifting generation from higher emitting carbon sources to lower

emitting carbon sources. But EPA failed to recognize the numerous technical issues that inhibit

the re-dispatch of coal-fired generation to gas-fired generation in Arizona. APS understands that

EPA claims it is not “requiring” the re-dispatch of coal-fired generation to gas-fired generation,



P a g e | 36

but any belief that Arizona would be able to meet the proposed carbon emission goals without a

large re-dispatch of coal-fired generation to NGCCs is simply credulous.

While some re-dispatch of coal-fired generation to NGCCs in Arizona is possible, the re-dispatch

of all of the state’s coal-fired generation to NGCCs creates a myriad of technical problems that

impacts the electrical grid reliability and energy cost. Such issues include the inability to meet

load demands during summer peak months, overloading of the electrical transmission system,

and the supply of natural gas necessary to reliably operate the NGCCs.

Based on EPA’s methodology, the largest carbon emission rate reduction for Arizona is based on

the re-dispatch of coal-fired generation to NGCC. APS analyzed the re-dispatch of coal-fired

generation to NGCC in Arizona as envisioned by EPA and identified numerous flawed

assumptions.

A. NGCC Generation Capacity Isn’t Available to Meet Peak Demands

A preponderance of the carbon emission rate reduction is based on the re-dispatch of coal-fired

generation to NGCC units located within the state. In fact, the proposed EPA goals are based on

100 percent of the coal-fired generation within Arizona being re-dispatched to NGCC units.

When evaluating the re-dispatch of coal-fired generation to NGCC in Arizona, EPA compared

the total 2012 actual NGCC generation output in the state to the potential annual generation

output to determine an annual capacity factor for NGCC. This is an important step for re-

dispatch because EPA has concluded that NGCC can reliably achieve a 70 percent capacity

factor and capped the re-dispatch to NGCC at this value. EPA determined the annual capacity

factor for NGCCs located in Arizona during 2012 was 27 percent. The difference between the

27 percent capacity factor calculated for 2012 and the 70 percent capacity factor cap allowed

EPA to conclude that all the existing coal-fired generation within Arizona could be re-dispatched

to NGCC units. Based on EPA’s calculations, when all coal-fired generation is re-dispatched to

NGCC units, the annual capacity factor for NGCC units located within the state increases to 53



P a g e | 37

percent. Therefore, EPA determined this generation re-dispatch was a viable option for setting

Arizona’s carbon emission rate goals and can be implemented commencing in 2020.

There are, however, a number of flaws in the assumptions used by EPA in its re-dispatch

analysis. When determining the potential generation capacity of the NGCCs located within

Arizona, EPA used the nameplate rating of the units; although, the nameplate rating of NGCC is

not the same as the net available output from the units. Net available capacity output is

influenced by a number of factors, such as site elevation, humidity, ambient temperatures,

turbine rating, and plant loads. In Arizona, peak electrical demand is typically proportional to

peak ambient temperatures. When electrical demand is at its highest is the same time when the

units’ net capacities are most limited due to ambient conditions. Because EPA did not take this

into consideration, NGCC generation capacity assumed by EPA is approximately 2,000 MW

higher than what is actually available during peak demand periods.

Table 1 below shows the difference between the generation capacities of NGCC units located

within Arizona assumed by EPA compared to the actual available capacities of these units.

Table 1

Generation Capacities of Arizona-Based NGCCs

NGCC Units Nameplate Summer Winter

MW MW MW

West Phoenix CC 1-3 396 255 276 West Phoenix CC 4 136 107 120 West Phoenix CC 5 570 490 506 Redhawk CC 1-2 1,140 934 1,007 Gila River CC 1 619 515 553 Gila River CC 2 619 515 553 Gila River CC 3 619 515 553 Gila River CC 4 619 515 553 Arlington CC 713 579 579



P a g e | 38

EPA’s second flawed assumption is the use of an annual capacity factor to determine the margin

of additional energy output that can be generated by NGCC in Arizona. In doing this, EPA must

have assumed the annual capacity factor for NGCC in Arizona is a rather flat line, when in

reality there is huge difference between the electrical demands in the summer months versus the

remainder of the year. For most years, the average summer demand is more than twice the

average demand for the remainder of the year.

In Arizona, the most critical period for utilization of generation capacity is the period from June

through September. APS used data from EPA’s Clean Air Markets Division to plot the historical

output from all of NGCC units located in Arizona during August 2012 to show the actual NGCC

capacity that is available during summer months versus how much the EPA assumes is available

for re-dispatch. Figure 2 below shows that during hour 16 on August 7, 2012, Arizona NGCCs

were generating 8,455 MW (net), and the coal and gas steam units were generating 4,098 MW

(net).

Santan CC 1,326 1,227 1,339 Kyrene CC 292 254 277 Desert Basin CC 646 577 625 Mesquite CC 1 692 536 594 Mesquite CC 2 692 538 588 Apache 82 72 72 Yuma Cogeneration Associates 63 52 54 Griffith Energy LLC 654 570 570 Harquahala CC 1-3 1,325 1,054 1,128

Total 11,202 9,305 9,947



P a g e | 39

Figure 2

In order to re-dispatch all coal and gas steam with NGCC generation as assumed by EPA, the

NGCCs would have to be operating at 12,533 MW (8,455 MW + 4,098 MW). The maximum

capacity assuming all units were fully available, however, was only 9,305 MW, a difference of

3,248 MW. Thus, only 850 MW of the 4,098 MW of required capacity would have been

available for re-dispatch in this hour, leaving 3,248 MW of demand that would still need to be

met. This suggests that 2,354 MW of coal would have been needed for reliability purposes. In

fact, there was not enough combined cycle capacity available for re-dispatch in the state of

Arizona for thirty of the thirty-one days in August, 2012 as indicated in Figure 3 below.



P a g e | 40

Figure 3

Figure 3 shows the results of the hourly analysis of the August 2012 data. Actual hourly

generation values for the Arizona NGCC units are indicated by the green line. APS then plotted

the additional generation that would be required from the NGCC fleet due to the re-dispatch of

coal-fired generation during the same period. This additional generation is indicated by the red

line. Figure 3 also shows the maximum capacity available from the NGCC units (horizontal

line). As seen in Figure 3 (red dashed line above horizontal line), almost 50 percent of the time

during August 2012 the energy demand that would normally be provided by coal-fired

generation exceeds the available capacity of all of the NGCC units in Arizona. Therefore, it is

not possible to re-dispatch all coal-fired generation to the existing NGCC fleet and meet

Arizona’s load demands with NGCC during peak demand periods.

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

1 46 91 136 181 226 271 316 361 406 451 496 541 586 631 676 721

MW

Hours

Arizona NGCC (9305 MW)Generation

August 2012

Above 100% NGCC Historical NGCC Re-Dispatched

527 GWHabove

100% CF

100% CF

64% Avg CF



P a g e | 41

The same analysis was performed for all hours of 2012 and is summarized in Table 2. Table 2

shows that for nearly 800 hours throughout the year, the NGCC units in Arizona could not meet

the energy demand.

Table 2

Excess Demand

Month GWH Hours January 0 0 February 0 0 March 0 0 April 0 0 May 3,147 6 June 138,369 145 July 212,637 235

August 526,832 361 September 51,529 92

October 20,313 22 November 0 0 December 0 0

Figure 4 below illustrates this shortage.



P a g e | 42

Figure 4

The green line in Figure 4 represents the actual capacity factor during the month of August 2012

for NGCC units in Arizona. The red line shows the increase in the capacity factor of the NGCC

units with the re-dispatch of the state’s coal generation. The re-dispatch capacity factor shows

over 360 hours (nearly 50 percent of the time) during the month where the demanded generation

exceeded the available generation. During this period, the average capacity factor for all the

NGCC units in Arizona would have to increase from 64 percent to 98 percent, which far exceeds

the 70 percent cap proposed by EPA. Because of the substantial increase in electrical demand in

Arizona during peak demands, the annual average capacity factor of 70 percent for the NGCC

units cannot be used as a basis for determining the additional capacity the NGCC units can

supply during peak demand periods.

0%

20%

40%

60%

80%

100%

120%

140%

160%

1 46 91 136 181 226 271 316 361 406 451 496 541 586 631 676 721

MW

H

Hours

Arizona NGCC (9305 MW)Capacity Factors

August 2012

Above 100% NGCC Historical NGCC Re-Dispatched

CF above 100%for 360 hours

98% Avg CF

64% Avg CF



P a g e | 43

Also, NGCC units are complex mechanical systems that malfunction even under the best of

readiness and preventive maintenance programs. It is naïve and unrealistic to assume there will

not be forced outages due to mechanical issues with these units from time to time. Such outages

will also reduce the availability of NGCC capacity, which is part of the reason why reserve

margins are required.

EPA suggested that the state’s utilities could just operate coal-fired generation during peak

demand periods, but this is not a viable option. The coal-fired EGUs in Arizona are large,

complex units. Typically, such units are not designed and engineered to sit idle for extended

periods of time. Such practices would challenge reliable operation of the units. Moreover, the

staffing and maintenance necessary to support such a scenario would not be economically

justified.

B. Re-Dispatch Will Require Construction of New NGCC Units

In addition to the shortage of electrical capacity to meet Arizona needs during peak demand

periods, there will also be a general shortage of electrical capacity if all of the state’s coal-fired

generation shuts down. APS, in conjunction with other Arizona utilities, commissioned PACE

Global to conduct a study of the impact of the Proposed Rule in Arizona. A copy of this report is

included with these comments.74 PACE Global used an AuroraXMP platform to model an

hourly chronological dispatch model to simulate the economic dispatch of power within a

competitive framework. Based on this modeling, PACE Global concluded that an additional 3.5

GW of new natural gas-fired capacity will be needed by 2020 to meet the load and reserve

margins. Of this additional generation, 2.3 GW is directly attributable to the Proposed Rule.75

Because utilities cannot be expected to prematurely make such an investment, the time frame for

designing, engineering, permitting, and constructing this new NGCC capacity is only two to

three years. It is unrealistic to believe that such a feat is possible. In addition to mammoth 74 PACE Global, Assessment of the Clean Power Plan (Nov. 2014). 75 Id.



P a g e | 44

amount of work that would be involved for the utilities, there is a huge concern regarding the

ability of original equipment manufacturers to supply the necessary equipment and the

availability of the labor necessary to construct the equipment in such as short time period.

C. Transmission System Cannot Support Re-Dispatch

The electrical high voltage transmission system in Arizona has evolved over time and, in part, is

designed to deliver power from remotely located power plants to the load. The Phoenix area is

by far the largest load center in Arizona, representing about 62 percent of Arizona load. It is

served by a complex network of generation and transmission primarily owned and operated by

APS and the Salt River Project Agricultural Improvement and Power District (“SRP”).

Figure 5

Arizona’s Existing Coal and NGCC Units and Transmission Infrastructure

Existing Coal Plants

Existing Gas CC Plants

345 kV and above Transmission System

Navajo generating station, excluded from AZ compliance



P a g e | 45

As shown in Figure 5 above, Arizona’s coal-fired units are predominantly located to the

northeast of the Phoenix load pocket while most of the NGCC units are located to the west and

southwest of Phoenix. The re-dispatch of all of Arizona’s coal-fired generation would change

the way power flows and impact a significant portion of Arizona’s transmission system. This

raises reliability concerns about the operation of the electrical transmission system.

One of the key measures of reliability of serving the Phoenix load pocket is Maximum Load

Serving Capability (“MLSC”). This is the maximum amount of load that can be served

assuming that all generation located inside the load pocket is operating at full capacity, and the

transmission system is importing as much power as possible, without overloading any circuits.

Transmission and local generation resources are planned based on projected load and MLSC plus

a reserve margin to account for contingencies. Based on APS’s assessment, there currently is

enough existing and planned local generation and transmission to reliably serve the APS Phoenix

load pocket into the 2020s. Re-dispatch of the coal-fired EGUs, however, will reduce MLSC

and jeopardize APS’s ability to reliably serve the Phoenix load.

APS questioned the impacts to the regional transmission system if the generation from all coal-

fired EGUs in Arizona were re-dispatched to the NGCC units by 2020 as envisioned in EPA’s

Proposal. Both the impacts on MLSC and impacts under light load conditions were evaluated.

Using industry standard power flow tools and methods, the limiting elements and transmission

system upgrades that would be needed to restore the MLSC to its present state and to mitigate

high voltage levels occurring under light load conditions were identified.76

The analysis was conducted using the planned transmission configuration in 2020 (Table 3

below) and assuming the generation from the following coal-fired EGUs is re-dispatched to the

NGCCs located in the state.

76 General Electric’s PSLF software, version 18.



P a g e | 46

Table 3

Planned 2020 Transmission Configuration77

Facility Capacity

(in MW) Apache Steam Units (AZ) 350 Cholla Units 1-4 (AZ) 1,027 Coronado Units 1-2 (AZ) 765 Navajo Unit 2 (AZ/Navajo Reservation) 75078 Springerville Units 1-4 (AZ) 1,609 San Juan Units 1-2 (NM) 680 Escalante Generation (NM) 247 Reid Gardner Generation (NV) 553

Three cases were developed for the MLSC studies:

• BASE CASE - MLSC of the Phoenix load pocket with coal-fired generation on line

• NO COAL - MLSC of the Phoenix load pocket after coal-fired generation re-dispatch

• NO COAL WITH UPGRADES - MLSC of the Phoenix load pocket with coal-fired

generation re-dispatched, and overloads mitigated

The results of the MLSC study show that under the “No Coal Scenario,” the MLSC of the

Phoenix Valley load pocket is reduced by 1,338 MW (from 16,321 MW to 14,983 MW). The

peak load plus margin value is 15,255 MW. Thus, this scenario results in a shortage of 272 MW.

77 Four Corners Units 1-3 (560MW), located on the Navajo Reservation in northwestern New Mexico, were included in EPA emission limit calculations, and were retired in 2013. According to EPA’s Proposal, EGUs located on Indian Country are not included in state goal calculations and are handled in EPA’s supplemental proposal for Indian Country & U.S. Territories. 78 Navajo Generating Station Unit 2 is expected to be shut down in 2019 due to reasons other than the Clean Power Plan.



P a g e | 47

Assuming the actual peak load does not exceed the projected peak load, and if all generators and

transmission lines are in service at the time of the peak, the load requirement is met and the

shortage occurs in the margin value. However, depending on what contingencies are occurring

at the time of the peak, it is possible that the load may not be met.

This analysis is a first cut at identifying the transmission impacts associated with the re-dispatch

of Arizona’s coal-fired EGUs. Due to the magnitude of the transmission changes that would be

necessary to resolve this situation, a more detailed engineering analysis would have to be

performed. Therefore, the cost and time required to implement the required upgrades has not

been confirmed. However, preliminary cost estimates are in the hundreds of millions of dollars

and the time to implement the transmission system upgrade will take years. As stated in

comments sent by the Southwest Power Pool to the Missouri Public Service Commission, “the

planning and construction of transmission upgrades can take up to eight and a half years and cost

up to approximately $2.3 million per mile for new 345 kilovolt transmission lines, excluding

substation costs.”79 Similarly, NERC explains that a construction timeline for a new high-

voltage transmission line can range from five to fifteen years, depending on the voltage class,

location, and availability of highly skilled construction crews.80

In addition, as highlighted by NERC, the current and planned natural gas pipeline infrastructure

in Arizona is inadequate for handling the increased gas demand due to the Proposed Rule. The

NERC Report notes the timing concerns because building additional pipeline infrastructure

investments can take three to five years to plan, permit, sign contract capacity, finance, and

build. The Proposed Rule’s timelines “would provide little time to add required pipeline or

related resource capacity by 2020,” according to NERC.81 APS also refers to the comments by

79 Responsive Comments of Southwest Power Pool, Inc., In the Matter of an Investigation of the Cost to Missouri’s Electric Utilities Resulting from Compliance with Federal Environmental Regulations, Missouri Public Service Commission File No. EW-2012-0065, at 3 (Sept. 16, 2014). 80 NERC Report at 20. 81 Id. at 10.



P a g e | 48

FERC Commissioners, as summarized in UARG’s comments, which show that EPA

inadequately accounted for economic and physical limitations on capacity expansions and

transmission. Based on recent APS experience, the lead time to permit, design, and construct

new transmission lines in Arizona is approximately 10 to 12 years.82

D. The Natural Gas Infrastructure is Inadequate

EPA examined the technical capability of the natural gas supply and delivery system to provide

increased quantities of natural gas to accommodate the shift in generation resources. EPA

concluded the natural gas supply system is adequate to achieve the proposed carbon emission

rate goals because the system already supports utilization capacity during peak periods and for

extended periods. Therefore, EPA concluded, it is reasonable to expect the system to support

utilization at less than peak periods and be reliable over the entire year. EPA also states that no

particular NGCC unit must achieve a specific utilization; therefore, if there are constraints, the

other NGCC units within the state could operate unconstrained. Lastly, EPA points to the ability

of transmission planners to repeatedly relieve bottlenecks and expand capacity. EPA states

“[t]here have been notable pipeline capacity expansion over the past five years, and substantial

additional pipeline expansions are currently under construction.”83

Similar to EPA’s approach with Subpart Da EGU heat rate efficiency and NGCC capacity

factors, EPA failed to recognize and analyze unique circumstances surrounding the issues. EPA

assumed that because the natural gas supply infrastructure has previously been adequate to

support peaks demands and for extended periods of time, that the system is adequate to continue

to do so into the future. This is simply not the case. EPA failed to consider the differences in the

gas supply infrastructure across the county. For example, the natural gas infrastructure is vastly

different in the West compared to the eastern infrastructure, as illustrated in Figure 6.

82 The assessment of the transmissions system’s short falls was independently confirmed by the PACE Global study. PACE Global concluded: “The lack of transmission import capacity limits load serving entities from displacing the retired units (coal) with the existing NGCC units.” 83 79 Fed. Reg. at 34,863-64.



P a g e | 49

Figure 6

U. S. Natural Gas Infrastructure

EPA also believes that since no particular NGCC unit is required to achieve a specific utilization

rate, that other NGCC units located in the state could make up for the loss due to a natural gas

supply issue. This does not apply well in Arizona because the state utilities are limited to one

primary gas supplier and, to a lesser extent, a secondary gas supplier. Furthermore, most all of

the NGCC units within the state are located within a small geographic region. Accordingly, any



P a g e | 50

natural gas supply interruption to any one particular NGCC is likely to also be an interruption to

most other NGCCs within the state.

Lastly, EPA relies on the ability of transmission planners to overcome any impediment and be

able to assure the natural gas supply is available to meet EPA’s proposed re-dispatch plan.

While transmission planners typically are quite skillful, the nation’s natural gas supply system is

in a tremendous state of flux and surely will have to overcome challenges never before

witnessed.

The PACE Global study identified that the incremental increase of natural gas demand resulting

from the increased utilization of NGCC units and the need for new NGCC capacity to fill the

void left from the retirement of coal-fired units will exceed the current natural gas supply

capacity.84 Arizona’s natural gas needs for power generation alone would increase by an

estimated 500 MMcf/d, representing more than a 55percent increase relative to Arizona’s current

power sector natural gas consumption. Natural gas consumption in Arizona would increase from

900 MMcf/d in 2015 to over 1,500 MMcf/d in 2020 and more than 2,500 MMcf/d by 2030. The

significant need for new pipeline infrastructure and capacity in a very short period of time will

further strain already stressed natural gas infrastructure in the state.

The already highly utilized and limited natural gas pipeline network serving Arizona, mainly El

Paso and Transwestern pipelines, would require expansion by the early 2020s to account for the

increased demand for natural gas by the power sector on top of expected demand growth in other

sectors. Based on modeling by PACE Global, the southern leg of the El Paso line is expected to

exceed its current capacity by 2022 and would require expansion by the early 2020s or sooner.85

The northern El Paso line would require expansion by 2025. The Transwestern pipeline, which

is already over 95percent utilized, would also require expansion to serve any additional future

demands.

84 PACE Global, Assessment of the Clean Power Plan, Section 3.1.4 (Nov. 2014). 85 Id., Section 3.5.2.



P a g e | 51

The development timeline to increase capacity is a minimum of 3 years with 4 or more years

being a more reasonable estimate. This means that the expansion would need to begin as soon as

Arizona’s implementation plan is finalized in 2017 to ensure reliable natural gas supply to the

state.

E. Re-dispatch in Arizona Will Lead to Significant Reliability Concerns

Much has been, and more can be, accomplished to reduce the carbon intensity in Arizona.

However, future actions must be rational and not lead to an unreliable or unaffordable energy

supply. Such an outcome could result in serious consequences in Arizona. Reliable and

affordable energy is a must in Arizona for the protection of human health. The unavailability of

energy for prolonged periods—either due to service interruptions or unaffordability—especially

during the hot summer months in Arizona could lead to devastating consequences.

1. Electric Supply Reliability

The reliability of the electrical supply can be affected by both an ability to supply the necessary

energy and an ability to deliver the energy. Compliance with the rule, as proposed, would result

in the elimination of all coal-fired generation in Arizona. As discussed above, EPA’s assumption

that coal-fired generation can be re-dispatched to NGCC units will not work as envisioned by

EPA, and incremental, lower-emitting generation resources cannot be permitted and brought

online in a two to three year period. Therefore, overall electrical supplies will be reduced and

will drive reserve margins lower. Even when accounting for planned new generation between

now and 2020, the reserve margins would remain around 15 percent meaning the loss of most or

all of Arizona’s coal-fired generation would require virtually a megawatt-for-megawatt

replacement of this generation to maintain safe reserve margins.



P a g e | 52

Figure 7

Arizona Reserve Margins with and without Incremental Natural Gas Builds

Source: PACE Global.

A functional and reliable transmission infrastructure is also needed to bring the new energy

supply to the load center. As previously discussed, new transmission will be needed for any new

NGCC units that must be built to replace the loss of coal-fired generation. EPA’s assumption

that coal-fired generation can be replaced by existing NGCC units is inaccurate, and the

magnitude and timing involved would make the state’s electric supply unreliable. Consistent

with the NERC Report, EPA should include in its final rule circumstances under which

compliance can be delayed to manage real time issues that will compromise electric reliability.



P a g e | 53

2. Natural Gas Supply Reliability

Arizona primarily has two long-haul natural gas pipelines, one owed by El Paso Natural Gas and

the other owned by Transwestern. Both of these natural gas pipelines are highly utilized to serve

demand in New Mexico, Arizona, and California and to meet rising exports to Mexico. Given

the current high capacity factor of natural gas pipeline usage in this region, the addition of CPP-

induced power generation demand in the Southwest could lead to a situation where the current

pipeline system is inadequate to serve all demand during periods of peak usage. In fact, NERC

found that “current and planned pipeline infrastructures in Arizona and Nevada are inadequate

for handling increased natural gas demand due to the [proposed CPP].”86 PACE Global’s natural

gas pipeline flow model analysis indicates that under the CPP, expansions would be needed to

meet consumption needs and maintain reliability on both the northern and southern legs of the El

Paso pipeline and on the Transwestern pipeline.

F. Utilities Do Not Control Merchant Plants

When EPA contemplated the re-dispatch of all the state’s coal-fired generation to NGCC, EPA

must have incorrectly assumed a command and control authority of the state utilities over the

merchant plants within Arizona. There are currently five merchant power plants located within

Arizona that control over 6,550 megawatts of generation capacity. The reality is that the

merchant plants are free to establish power purchase agreements (PPAs) with whomever they

want. The energy from these units may or may not be available for use within Arizona. In fact,

some of the merchant plants already have long-term PPAs to supply their energy outside of

Arizona. The potential unavailability of this generation creates an additional deficit for meeting

Arizona’s energy needs. The most reasonable choice to overcome this deficit is to use some of

the existing coal-fired generation within Arizona.

As noted above, Arizona’s proposed final goal is 702 lb/MWh. When running in their optimum

condition and under prime operating scenarios, NGCC units operate at a much higher carbon 86 NERC Report at 10.



P a g e | 54

emission rate than 702 lb/MWh. (Even EPA acknowledges in its Section 111(b) rule for new

units that a new NGCC unit cannot be expected to have less than 1,000 lb/MWh.) Thus, a likely

scenario is that an existing merchant-owned NGCC unit operating under prime conditions would

emit approximately 198 lb CO2/MWh more than the Arizona goal. To complicate matters, this

same merchant generator could be running this EGU for the purposes of selling the energy

outside of the state of Arizona. Thus, under the Proposed Rule, which requires carbon emissions

from EGUs to be counted in the state in which they are generated, Arizona would have to

account for these merchant emissions when evaluating compliance with its 111(d) plan, making

it unduly difficult for Arizona utilities to comply. Moreover, under this scenario, Arizona would

be penalized for carbon emissions without reaping the benefit of having re-dispatched a coal-

fired EGU emitting an estimated 2,200+ lb CO2/MWh to a lower-emitting NGCC unit. In this

scenario, the state would have to add the carbon emissions from the coal-fired EGU to serve its

domestic load to the carbon emissions from any NGCC units generating electricity for out-of-

state consumption to calculate compliance. APS recommends that the EPA include a provision

in the final rule to account for such interstate transfers such that the generator’s home state is not

harmed.

XIII. RE-DISPATCH IN ARIZONA WILL UNREASONABLY INCREASE ENERGY COSTS

Energy costs for Arizona are expected to increase disproportionately when compared to other

states. Arizona has one of the most severe carbon emission reduction targets of all states with a

reduction below its baseline of 52 percent. Most states’ targets are well below 50 percent and

some are less than 20 percent (see Figure 7 below).



P a g e | 55

Figure 8

Reduction of 2012 CO2 Emissions Required by Clean Power Plan Final Goals

Source: EPA.

Other states like Washington and South Carolina also have a large percentage reduction,

although the effects are much different. Washington has only a small amount of coal-fired

generation, so from a percentage standpoint, there is a large reduction but not in terms of actual

generation-switching measures. South Carolina is impacted not by building block 2, as it has

limited existing NGCC capacity, but rather by building block 3 due to the proposed new nuclear

generation capacity in the state. Arizona stands out as the state most impacted by building block

2 due to its high coal-fired generation and significant existing NGCC capacity.

As a result of the overly stringent carbon emissions reduction targets, there are significant cost

impacts to Arizona due to new infrastructure investments, including power and natural gas

infrastructure, operational changes to the generation fleet, and the recovery of stranded

investments. A disproportionate increase in energy cost places Arizona at a significant

competitive disadvantage compared to neighboring states. As states compete to attract new

<20%20 - 30%30 - 40%40 - 50%>50%

2030 Target Reduction from 2012 Baseline

Washington3.7 TWh impacted coal gen.3,485 MW existing NGCC

Arizona24.3 TWh impacted coal gen.11,202 MW existing NGCC

South Carolina28.4 TWh impacted coal gen.2,839 MW existing NGCC



P a g e | 56

businesses to increase state revenues and job opportunities, the cost of energy to operate the

business plays a large role in the decision as to where to locate the business. If energy costs

increase disproportionally in Arizona, as compared to competing states, Arizona will lose

opportunities to attract new businesses, which will limit growth within the state. This is a

consequence of the Proposed Rule EPA presumably did not intend.

A. Cost of New Generation Infrastructure

PACE Global analyzed the cost differential between installations of new NGCC capacity under

the Proposed Rule as compared to a more level glide path approach for Arizona. Achieving the

necessary new generation capacity that will be necessary to comply with the Proposed Rule

increases the cost of compliance by $1.9 billion.87

Table 4

Capital Costs for New Natural Gas Generation by Scenario (2013$M)

Total 2020 Cumulative Total 2020 – 2030

2020 Cost Attributed to

CPP

Cumulative 2020 – 2030 Cost Attributed

to CPP EPA Building

Blocks $3,000 $8,067 $1,991 $1,900

Arizona Glide Path $3,000 $6,168 $0 $0

Source: EPA, PACE Global.

In addition to the cost of new generation capacity, significant transmission investment could also

be required depending on the location of the new capacity versus the load centers in the state.

B. Economic Assessment Including the Cost of Stranded Assets

To meet the interim and final goals of the Proposed Rule, some or all coal-fired generation in

Arizona would need to be retired. Arizona has a relatively new coal-fired fleet and several of the

coal-fired units recently have been equipped with expensive pollution control technology. The

87 Assessment of the Clean Power Plan, Section 5.1, PACE Global (Nov. 21, 2014).



P a g e | 57

value of the assets that could become stranded in 2020 is $3.8 billion.88 The impact of this

stranded investment is exacerbated by the fact that new generation would have to be built to

replace the stranded assets, thus compounding the cost impact to Arizona customers.

C. Cost of Changing Coal Plant Operational Behavior

As a result of the Proposed Rule, there is a potential to change the operational behavior of any

coal-fired generation that may remain operational within Arizona. Coal-fired units may be

required to have increased startup/shutdown cycles and load following cycles. While each of

these cycles causes some measure of wear and tear in excess of steady state operation, cold starts

are generally the most damaging and load following the least damaging. Load cycling may also

increase equipment wear and therefore cost. These impacts generally manifest themselves in

some combination of unit de-rating due to equipment damage, increased capital and maintenance

costs due to repairing prematurely worn parts, and increased fuel consumption either from

increased startups and/or lower heat rates resulting from part load operations.

Depending upon the cycling type, from 49 to 62 percent of the cycling cost results from

increased capital and maintenance costs associated with increased maintenance frequency,

inspections, and repairs. Moreover, 23 to 29 percent of the cost increase results from forced

outages requiring not only repairs (parts and labor), but also purchase of replacement power.89

D. Cost of New Natural Gas Infrastructure

According to a 2013 U.S. pipeline economics study conducted by Oil and Gas Journal, new gas

pipeline construction costs have averaged $155,000 per inch-mile. For smaller pipelines, less

than 12 inches in diameter, costs are assumed to range from $20,000 to $70,000 per inch-mile.

Adding capacity potentially can cost less if the design capacity of the pipeline allows for

incremental gas compression. However, the recent demand for new natural gas pipeline capacity

88 Id., Section 5.4.1. 89 Id., Section 5.5.



P a g e | 58

is driving costs even higher. Because of the limited number of companies that install new

natural gas pipelines, the cost and time for installation is escalating as demand continues to grow.

The northern leg of the El Paso Natural Gas pipeline is likely to become highly constrained by

2020 and in need of expansion. At 330 miles in length, new transmission capacity on this

segment may require hundreds of millions of capital expenditures. Permitting, siting, and

construction timelines for gas pipelines can take three to five or more years, highlighting the

need incumbent upon gas-fired generators to plan for adequate transmission capacity.

E. Cost of Natural Gas

Increases in natural gas demand resulting from implementation of the CPP are likely to raise gas

prices and overall consumer prices nationally and in Arizona. The impact of additional gas

demand required for compliance would come at a time when non-power sectors and exports

(liquefied natural gas and pipeline exports to Mexico) are also growing rapidly. PACE Global

estimates that all other things being held constant, power sector consumption could increase over

18 Bcf/d by 2030 over a business-as-usual (BAU) projection of natural gas demand nationwide.

Figure 9

Projected National Natural Gas Power Sector Demand (Clean Power Plan v. BAU)

05,000

10,00015,00020,00025,00030,00035,00040,00045,00050,000

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

MMcf/

d

US Demand

CPP Building Block Case BAU Case

U.S. Power Sector Gas Demand – CPP Building Block Case v. BAU Case



P a g e | 59


PACE Global estimates that this incremental increase in national consumption levels would

result in Henry Hub price increases close to $3/MMBtu over a non-carbon scenario.

Figure 10

Projected Henry Hub Natural Gas Pricing (Clean Power Plan v. BAU)


In addition, the market has recently seen a sharp increase in the volatility of market prices in

regions where shale gas development has outstripped the capability of existing infrastructure to

deliver natural gas to markets where the gas is needed. Price spikes in the Northeast last winter

exceeded $123.50/MMBtu. As demand increases rapidly, this issue could become a much more

widespread problem than in just the Northeast. In fact, we have already seen evidence of market

volatility in the Northeast because infrastructure has not kept up with demand (and available

$0.00

$1.00

$2.00

$3.00

$4.00

$5.00

$6.00

$7.00

$8.00

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

Rea

l 201

3$/M

MB

tu

Henry Hub

CPP Building Block Case BAU Case

Henry Hub Pricing – CPP Building Block Case v. BAU Case



P a g e | 60

supplies). Volatility could be very significant with growth in power sector natural gas demand

projected under the CPP.

XIV. APS’S PROPOSED SOLUTION TO ADDRESS ISSUES ASSOCIATED WITH RE-DISPATCH

The re-dispatch of coal-fired generation to NGCC units must be the center of any attempt by the

utilities to achieve the proposed interim and final carbon emission rate goals set forth for

Arizona. The preponderance of the carbon emission rate reduction in Arizona is based on this

re-dispatch. Because all coal-fired generation in Arizona is presumed to be re-dispatched to the

NGCC fleet, emission rate reductions from unit efficiency improvements are meaningless. RE

and EE programs are strong, and will continue to be strong in Arizona; however, there is only so

much that can be gained through RE and EE programs.

APS has clearly shown there are significant issues associated with the re-dispatch of coal-fired

generation to NGCC units in Arizona. The most significant of these issues includes insufficient

capacity to meet Arizona peak energy demands, instability of the electrical transmission system

due to changes in energy flow through the system, and concerns around the availability and

pricing of natural gas. Resolution of these issues is unachievable by 2020, when virtually all of

Arizona’s coal generation would have to be re-dispatched to meet the first interim goal. Even

with more time, the final emission rate goal proposed by EPA leads to an irrational position. It is

certain any attempt to achieve the carbon emission rate goals proposed for Arizona will lead to

an unreliable electrical supply system, and higher energy cost for Arizona.

APS urges EPA to reconsider the proposed carbon emission rate goals in light of all the issues

associated with re-dispatch that have been identified. Clearly, the interim goals are unrealistic

and unattainable for Arizona. Furthermore, the final Arizona goal must be revised to be more

realistic and to allow for a sensible energy mix in Arizona. This outcome will result in the best

solution to balance environmental protection with the need for reliable and affordable energy in

Arizona.



P a g e | 61

APS is encouraged by EPA’s acknowledgment in the NODA of utility concerns that the

Proposed Rule fails to give owners the opportunity “to fully take advantage of the remaining

asset value of existing coal-fired generation.”90 EPA attempts to address this concern through

introduction of the book life concept.91 APS supports EPA’s acknowledgement that building

block 2 may be inadequate to address industry concerns regarding stranded investments.

However, it is important to stress that book life is not the same as remaining useful life because

while the former is a depreciation concept that is used in accounting and ratemaking, the latter

necessarily describes the asset’s remaining useful life. Thus, while a set number of years (e.g.,

40 years for unit life or 20 years for major pollution control retrofit projects) may be appropriate

for purposes of depreciating an asset, such an approach would be wholly inappropriate for

considering an asset’s remaining useful life, which must be determined on a case-by-case basis.

The CAA requires that “[i]n promulgating a standard of performance under a [111(d)] plan . . .

EPA shall take into consideration, among other factors, remaining useful lives of the sources in

the category of sources to which such standard applies.”92 Accordingly, EPA does not have the

legal authority to ignore the remaining useful lives of affected sources when setting a state’s

goals or to replace consideration of this factor with a concept the Agency might prefer.93 The

CAA further requires that EPA’s regulations “permit the State in a applying a standard of

performance to any particular source under a [111(d)] plan . . . to take into consideration the

remaining useful life of the existing source to which such standard applies.”94 EPA may not by

rule eviscerate this statutory obligation under the guise of its illusory notion of state “flexibility.”

As discussed above, because the state of Arizona must re-dispatch all of its coal-fired generation

90 79 Fed. Reg. at 64,544. 91 79 Fed. Reg. at 64,549. 92 40 U.S.C. § 7411(d). 93 See 79 Fed. Reg. at 34,925 (asserting “flexibility provided in the state plan development process adequately allows for consideration of the remaining useful life of the affected facilities and other source-specific factors and, therefore, that separate application of the remaining useful life provision by states in the course of developing and implementing their CAA section 111(d) plans is unnecessary). 94 40 U.S.C. § 7411(d)(1)(B).



P a g e | 62

under building block 2 in order to comply with the state’s interim goal, the state actually has no

flexibility to adequately consider the remaining useful lives of Arizona’s coal-fired fleet, as EPA

suggest in the Proposed Rule. The Agency’s assertions of flexibility ring quite hollow when one

considers the fact that EPA’s application of its building blocks and BSER to Arizona leaves

insufficient room for the state to make any accommodations whatsoever to consider the

remaining useful life factor.

Although, for the foregoing reasons, APS believes it is necessary for EPA to consider each

affected facility’s remaining useful life when setting the state goals, APS also believes it would

be appropriate to adjust Arizona’s interim and final goals to account for the book life of

generation assets and major pollution control retrofits, as well as commitments made by utilities

to shut down or convert coal-fired units to natural gas during the interim period. Consistent

therewith, APS proposes the following few targeted changes to the final rule:

1. For purposes of goal setting under Building Block 2:

a. Re-dispatch from coal-fired EGUs to natural gas combined cycle (NGCC) EGUs should

occur upon the later of any of the following, if re-dispatch would occur prior to January

1, 2030:

i. January 1, 2020;

ii. 40 years after initial commencement of operation; or

iii. 20 years after commencement of operation of major pollution control retrofits,

such as selective catalytic reduction (SCR), flue gas desulfurization (FGD), or

baghouses at any EGU if installation occurred prior to issuance of the final

111(d) rule, or after commencement of operation of selective non-catalytic

reduction (SNCR) or electrostatic precipitators (ESPs) at an EGU owned by a

small utility as defined by the Federal Energy Regulatory Commission



P a g e | 63

(FERC) if installation occurred prior to the first year of the compliance period

(i.e., 2020).

b. For coal-fired EGUs that either shut down or convert to natural gas-fired operation, re-

dispatch would occur as specified in an applicable implementation plan or enforceable

Title V permit, provided that such commitment is entered prior to the effective date of the

final rule and the date of shutdown or natural gas conversion is prior to January 1, 2030.

c. Coal-fired EGUs that do not re-dispatch prior to January 1, 2030 under paragraphs 1.a or

1.b remain coal-fired EGUs for purposes of calculating the interim and final Goals.

2. For purposes of goal setting, when re-dispatching to NGCC, a rate of 1,000 lb CO2/MWh

should be used, consistent with the most stringent standard in the EPA’s proposed New

Source Performance Standard for EGUs.95

3. The State should establish the Interim Goal in its State Plan based upon EPA’s building

block approach as modified by paragraphs 1 and 2 above.

As further explained in the comments of the Arizona Utilities Group (of which APS is a

member) filed contemporaneously with these comments, incorporating these recommendations

into the final rule would address the primary technical problems associated with the current

proposed Arizona goals, result in a final rule that would not threaten electric reliability, provide

substantial reductions in carbon emissions both in Arizona and nationwide, and be consistent

with the requirements of CAA Section 111(d).

XV. EPA SHOULD NOT ADOPT THE “MINIMUM LEVEL OF GENERATION SHIFT” CONCEPT

Additional approaches to Building Block 2 that EPA describes in the October 2014

NODA present numerous achievability challenges. EPA should not expand the types of natural 95 79 Fed. Reg. 1,433.



P a g e | 64

gas facilities in the BSER—to include new NGCC generation and co-firing with natural gas—

due to many of the same feasibility concerns described above. Furthermore, APS opposes the

inclusion of co-firing natural gas at existing coal plants as part of building block 2 because of the

technical obstacles to co-firing at many coal plants. Of the four coal-fired units at the Cholla

Power Plant, only one unit currently has the ability to combust a small amount of natural gas for

startup and flame stabilization of the unit. The maximum current amount of heat input from

natural gas is less than one percent. The other Cholla units use fuel oil for startup and flame

stabilization. The two coal-fired units at Four Corners are equipped with the ability to combust a

small amount of natural gas for startup operations and early flame stabilization. However, the

amount of heat input from natural gas combustion once again is limited to a small fraction of the

total heat input to the units. An enormous capital investment to modify the boilers and increase

the natural gas supply infrastructure, which would also take numerous years to complete, would

be necessary to achieve a material reduction in carbon emissions from co-firing natural gas.

Such actions simply cannot be rationalized or cost justified.

EPA should not incorporate a “minimum level of generation shift”96 from higher-emitting to

lower-emitting sources into its approach to building block 2 for states with fossil steam

generation, nor should EPA apply building block 2 on a regional basis. Any such minimum level

of generation shift would be an inflexible requirement. Moreover, EPA has not determined that

either a minimum level of generation shift or a regional calculation of increased dispatch of

NGCC capacity is adequately demonstrated. Such a determination is not supportable, since

individual states have limited existing NGCC capacity and there are constraints on building new

capacity.

96 79 Fed. Reg. at 64,549-50.



P a g e | 65

XVI. APS’S CONCERNS WITH BUILDING BLOCK 3 CONCEPTS

A. EPA Should Mandate Bundling Of Renewable Energy Credits

As described in Section I of these comments, the components of building block 3 are not within

the realm of BSER for establishing carbon performance standards for individual EGUs.

However, APS recognizes the importance of renewable energy and the importance it plays in

meeting Arizona’s energy demand. APS has actively been engaged in interconnecting renewable

resources to its grid for decades and has been very successful in achieving the standards

established by the Arizona Corporation Commission (“ACC”) since inception of the state’s

Renewable Energy Standard (“RES”).97

APS has concerns with how the contribution of renewable energy will be credited towards

compliance with the state goals, especially in terms of renewable energy credits (RECs). APS

believes that only bundled RECs should be counted towards meeting the state carbon emission

rate goals. Bundled RECs are a requirement of the ACC-mandated RES and should be required

under EPA’s 111(d) final rule. There could be an adverse effect in Arizona and other states if

unbundled energy RECs are allowed to be counted towards a state’s emission rate goals.

Allowing an entity to count the environmental attributes of renewable energy without taking the

energy itself may increase the build out of renewable capacity in states with the capacity and

weather conditions to support robust wind or solar resources. Because utilities must purchase the

energy produced from those intermittent resources pursuant to Public Utility Regulatory Policy

Act rules, utilities would have to manage the integration of that energy onto their systems.98 The

benefit of RECs would be retained by the owner of the systems, but the associated integration

costs could be passed on to the local ratepayers.

Specific to the Arizona RES, a minimum portion of the renewable energy requirement must be

met through Distributed Generation (DG), and this has an effect on the amount of utility-scale

97 The ACC established the RES in 2007. A.A.C. R-14-2-1801 et. seq. (eff. Aug. 14, 2007). 98 16 U.S.C. 824(a)(3) (enacted Nov. 9. 1978).



P a g e | 66

capacity that is approved for utilities. Therefore, EPA should allow all electric generation and

displacement resulting from renewable DG technologies that occur within a state to be fully

applicable towards the state’s 111(d) compliance. Depending on the state and local utility rate

structure, much of the growth in a state’s RE capacity may come from distributed systems. In

particular, Arizona is among the highest growth states for both the total number of PV systems

and total solar capacity installed.99

B. At-Risk Nuclear Should Not Be Included In Goal-Setting

APS is the operator and partial owner of the Palo Verde Nuclear Generating Stations (PVNGS).

PVNGS is the largest operating nuclear power station in the United States. As an operator and

owner of a nuclear power plant, APS commends EPA for recognizing the importance nuclear

power plays in lowering and avoiding carbon emissions. APS also appreciates EPA’s attempt to

preserve “at-risk” nuclear power. Based on an EIA report, EPA has concluded that

approximately six percent of nuclear capacity is “at-risk” of shutdown due to continued

economic challenges. Therefore, EPA proposed to allow approximately six percent of the

nuclear capacity within any state in which nuclear plants are located to be counted towards

achievement of the state emission rate goals.

However, EPA also used the “at-risk” nuclear to set the state’s emission rate goals. Therefore,

nothing is really gained from use of this carbon-free source of generation. All existing nuclear

power plants will eventually close. When this occurs, states will have to find ways to replace a

large amount of zero-carbon generation with other zero-carbon generation or be at risk of being

in non-compliance with meeting their 111(d) goals. Because there currently are no cost-effective

mechanisms to achieve this, the “at-risk” nuclear generation should not be included in states’

carbon emission rate goals. The six percent of “at-risk” nuclear generation should be eliminated

99 In 2013, Arizona ranked 2nd in total solar energy installed in the U.S. with 1,899 MW; http://www/seia.org/state-solar-policy/arizona.



P a g e | 67

from the final rule, and for states with nuclear generation, the state goals should be adjusted

upward by approximately six percent.

EPA clearly recognizes the importance that nuclear generation has on reducing and avoiding

carbon emissions. In the Proposal, EPA stated that “[i]ncreasing the nuclear capability relative

to the amount that would otherwise be available to operate is therefore a technically viable

approach to support reducing carbon emissions from affected fossil fuel-fired EGUs.”100 APS is

in full agreement and offers the following as a means of ensuring nuclear generation is

incentivized to reduce carbon emission rates. While EPA states in the Proposal that it estimated

a 90 percent average utilization rate for U.S. nuclear units, APS has different information.101

Based on industry data, the average lifetime capacity factor of the 100 nuclear plants currently

operating is 81.6 percent.102 Of these, based on the three-year average between 2011to 2013, the

plants in the fourth quartile only operated at an average capacity factor of 70.8 percent.

Therefore, to encourage nuclear generation to operate at peak performance, EPA should allow

the energy from any nuclear power plant in excess of the 80.3 percent average lifetime capacity

factor for all current nuclear plants to be used for compliance with the states’ carbon emission

rate goals. But because this is meant to be an incentive for peak performance of the U.S. nuclear

fleet, none of the nuclear generation should be used to calculate the states’ goals. This nuclear

generation, like renewables, is a zero carbon-emitting generation resource. This approach adds

flexibility to the program and assists the states that have nuclear capability in achieving their

final carbon emission rate goals.

C. Renewable Energy Is Not a Suitable Substitute for Base Load Power

The alternative approach that the Agency presents in the NODA—to apply generation from

building blocks 3 and 4 to reduce fossil generation below 2012 levels in the goal calculations— 100 79 Fed. Reg. at 34,870. 101 Id. at 34,871. 102 http://www.nei.org/Knowledge-Center/Nuclear-Statistics/US-Nuclear-Power-Plants.



P a g e | 68

would only serve to exacerbate the infeasibility and inflexibility of the currently proposed

performance goals, and should therefore not be incorporated.103 While fossil fuel generation is a

steady source of base load power, renewable energy generation is intermittent and there are not

sufficient EE reduction opportunities to make up for loss of base load power, taking into account

the projected increase in electricity demand.

XVII. APS’S CONCERNS WITH BUILDING BLOCK 4 CONCEPTS

APS supports the use of EE as a contributor to setting statewide goals for the carbon emission

rate. APS has been an early adopter and strong supporter of EE programs for over ten years.

APS has already achieved over 2,600,000 MWh of cumulative first-year savings from EE

programs since 2005, resulting in over 2.3 billion pounds of avoided carbon emissions.

However, APS has a number of concerns with the Agency’s methodology in respect to building

block 4.

A. It is Unreasonable to Assume It is Feasible to Continuously Sustain a High Level of Savings from Energy Efficiency Programs

As the largest electric utility in the state of Arizona, which has been recognized by EPA to be a

“best practice” state for achieving energy savings from EE programs, APS disagrees with EPA’s

assumption of a continued high level of savings coming from EE in setting the carbon emission

goals for Arizona. Due to a variety of technological, economic, and market factors, EE savings

have a finite limit or maximum market potential that can be achieved. Not all energy

consumption can be eliminated through conservation or efficiency. As a state approaches this

finite limit on energy savings, less incremental savings can be achieved with each passing year.

States like Arizona that have been deploying EE for a long period of time and have already

achieved significant savings are closer to that finite limit than states that are just beginning to

implement EE. Therefore, APS is concerned about the assumption made in setting the carbon

103 79 Fed. Reg. at 64,552.



P a g e | 69

emissions goal that Arizona can sustain a level of savings from EE at 1.5 percent of sales for the

next 13 years (2017 to 2030). We believe that the rate of savings will decline over that period of

time due to the limit on the market potential for savings.

This position is also support by the analysis conducted by Pace Global, which identified that the

1.5 percent energy efficiency standard has only been demonstrated in top tier states for a single

year.104 Additionally, EPA only used one data source for the 1.5 percent per year standard. Pace

Global goes on to point out the next highest efficiency potential as reported by the Electric

Power Research Institute (EPRI) in 2014 is 0.6 percent per year.105 Accordingly, the EPA data

does not strongly support that a 1.5 percent energy efficiency standard can be achieved year over

year for multiple years, and Pace Global recommends that for the purposes of goal setting, EPA

should use the 0.6 percent per year standard identified in the EPRI study.

APS supports the 0.6 percent energy efficiency standard identified in the ERPI study and as

recommended by Pace Global for the purposed of goal setting. However, should EPA chose to

not lower the 1.5 percent requirement to a 0.6 requirement, at the very least APS recommends

that EPA use a scaled rate between 0.6 percent and 1.5 percent per year during the 2017 to 2030

time frame based on the length of time that each state has been engaged in EE programs.

The length of time that EE programs have been implemented affects states’ ability to achieve EE

improvements as they approach the end of the initial compliance term and into the future. States

would apply a reduced annual EE rate as a percent of retail sales, based on a pre-determined

scale, which would take into account the amount of historical achievements in EE savings. For

example, a state that has already achieved a significant portion of the total market potential for

EE savings would be required to achieve EE savings at a lower percent of sales than a state that

has not yet seen any significant savings from EE programs. The rationale for this is that those

utilities in states that have been successfully implementing EE programs prior to 2017 have

104 Assessment of the Clean Power Plan, Section 3.1, PACE Global (Nov. 21, 2014). 105 Id.



P a g e | 70

already taken advantage of the “low hanging fruit”—those EE measures that are less expensive

to achieve and may occur in larger numbers. These same utilities now have to be more creative

and spend more per unit of energy saved to convince the next wave of customers that they should

invest in EE. In plain language, proactive utilities will find it more difficult and more expensive

to maintain the same level of incremental EE savings that they have already achieved compared

to a utility that has not yet achieved that same level of savings.

B. EPA Should Defer to Established State Evaluation, Measurement, and Verification Protocols

The ACC adopted an Energy Efficiency Resource Standard (“EERS”) in 2010 that requires a

cumulative energy savings from demand-side management programs equivalent to 22 percent of

each affected utility’s prior year retail sales by 2020. APS has been in compliance with the

EERS since it took effect in 2011 and has significant energy savings accumulated prior to 2011.

These savings are allowed to count toward the ACC goal of 22 percent. In putting together a

comprehensive suite of EE programs over the last ten years, APS also already has in place

industry-standard protocols for Evaluation, Measurement, and Verification (“EM&V”) of

reported savings. These EM&V standards are also required by the EERS rules. APS contracts

with a nationally recognized, independent third-party firm to conduct the EM&V work and

report the verified energy savings results to the ACC. Because utilities in Arizona already have

accepted methods for calculating savings from EE, APS believes those same methods should be

used for counting savings toward Arizona’s emission rate goals. Savings from Arizona’s

demand-side management (“DSM”) programs meet EPA’s four criteria for being creditable in

that they are quantifiable, surplus, enforceable, and permanent.

1. Quantifiable

In order to report accurately on program savings to the ACC, APS has put in place

comprehensive procedures to track and measure the energy savings resulting from DSM

programs. These procedures follow industry best practices for quantifying and verifying savings

and include, without limitation, the following activities: tracking and verifying the number of



P a g e | 71

customers participating in APS’s DSM programs and the number of energy efficient units (e.g.,

light bulbs, HVAC systems, duct repair installations, etc.) that are installed; comparing the

average energy usage of a high-efficiency piece of equipment or appliance to a baseline of the

average energy use of a standard-efficiency piece of equipment or appliance that would have

been installed but for the DSM program to determine average annual energy savings resulting

from the program; and adjusting for energy savings that is occurring naturally among APS’s

customers or would have occurred without the DSM program. Post-implementation

measurement and evaluation of savings is done using a combination of field meter data

collection, energy bill history analysis, and customer surveys conducted from a sample of all

installations.

2. Surplus

Due to stricter requirements for energy usage written into appliance efficiency standards and new

construction building codes, energy usage per customer has been naturally declining over time.

That trend is often offset by the trends for new homes to be larger and all homes to contain more

electricity-using appliances and equipment than they did in the past. These “naturally occurring”

trends are captured and projected forward in APS’s econometric load forecast. APS carefully

designs and implements DSM programs to drive incremental energy savings. Where applicable,

programs in the APS portfolio are developed around national EPA/U.S. Department of Energy

ENERGY STAR requirements and other national standards to ensure they promote high levels of

efficiency. Incremental savings from APS’s DSM programs are considered to be additional to

the naturally occurring conservation efforts because the program savings would not have

occurred without the additional stimulus provided by APS’s DSM programs. In order to

estimate the net impact of DSM program participation on energy use, billing analysis of program

participants is done using experimental design principles, which control for extraneous variables

impacting energy usage.



P a g e | 72

3. Enforceable

Affected utilities’ compliance with the EERS is mandatory, and failure to comply could result in

financial loss and the imposition of other sanctions by the ACC. While APS cannot dictate

individual customer participation in a DSM program, APS can structure its programs to appeal to

a broad customer base and be a compelling value proposition to APS customers, thereby

encouraging their participation.

4. Permanent

Arizona’s EERS employs the principle of persistence. Specifically, the EERS provides that an

affected utility’s energy savings used to meet the standard will be assumed to continue through

the year 2020 or, if expiring before the year 2020, to be replaced with an EE program having at

least the same level of efficiency. Many EE measures in the APS portfolio, such as EE upgrades

for newly built homes and commercial buildings, as well as improvements to existing homes’

and facilities’ ductwork and insulation, are designed to last for a substantial part of the

structure’s lifetime. Across APS’s entire DSM portfolio of programs, the average amount of

time that an energy efficient measure is expected to last is approximately 10 years. Experience

in Arizona and in other states with similar EE standards suggests that after a period of 10 years, a

new EE measure installed will be at least as efficient as the old one being replaced (e.g., from

CFL bulbs to more efficient LED bulbs). Very seldom is that not the case and when it is, savings

estimates can be adjusted to reflect the migration away from EE for some small percentage of

customers.

Because Arizona's robust DSM program includes all of the elements required under the most

exacting pathway included in EPA's Roadmap—the control strategy pathway—and has an

established record of leading to meaningful, cost effective carbon reductions, consistent with the

primary role given to the states under Section 111(d) of the Act, the Agency should recognize

Arizona’s program as being “equivalent” and allow Arizona to use the EM&V requirements it

currently has in place. Deferring to Arizona to determine the standards for measuring EE and

giving the state broad flexibility concerning how to measure the energy savings resulting



P a g e | 73

therefrom would be consistent with the principle of cooperative federalism intended by Congress

when it enacted Section 111(d) of the Act.

C. EPA Should Allow States to Bank Savings from Energy Efficiency Programs and such Programs Should be Broadly Defined

APS believes the savings achieved to date should be banked and allowed to count toward

compliance as well as future savings. We also believe that the types of programs that are

allowed to count as EE under the CPP should be defined broadly and should be consistent with

the menu of programs that has been approved as cost-effective by the state jurisdictions. Finally,

the amount of EE done by a state will also be a function of the program and budget approval by

the state public utility commission (“PUC”) or the local jurisdiction. APS believes EPA should

specifically acknowledge the role of state PUCs in providing funding for EE programs.



Attachment

PACE Global Assessment of the Clean Power Plan

4401 Fair Lakes Court Fairfax, VA 22033 USA Phone: 1.703.818.9100 www.paceglobal.com

Setting the Pace in energy

Assessment of the Clean Power Plan

Prepared for:

The Arizona Utility Group

November 21, 2014

This Report was produced by Pace Global, a Siemens business (“Pace Global”) and is meant to be read as a whole and in conjunction with this disclaimer. Any use of this Report other than as a whole and in conjunction with this disclaimer is forbidden. Any use of this Report outside of its stated purpose without the prior written consent of Pace Global is forbidden. Except for its stated purpose, this Report may not be copied or distributed in whole or in part without Pace Global’s prior written consent.

This Report and the information and statements herein are based in whole or in part on information obtained various sources as of November 19, 2014. While Pace Global believes such information to be accurate, it makes no assurances, endorsements or warranties, express or implied, as to the validity, accuracy or completeness of any such information, any conclusions based thereon, or any methods disclosed in this Report. Pace Global assumes no responsibility for the results of any actions and inactions taken on the basis of this Report. By a party using, acting or relying on this Report, such party consents and agrees that Pace Global, its employees, directors, officers, contractors, advisors, members, affiliates, successors and agents shall have no liability with respect to such use, actions, inactions, or reliance.

This Report does contain some forward-looking opinions. Certain unanticipated factors could cause actual results to differ from the opinions contained herein. Forward-looking opinions are based on historical and/or current information that relate to future operations, strategies, financial results or other developments. Some of the unanticipated factors, among others, that could cause the actual results to differ include regulatory developments, technological changes, competitive conditions, new products, general economic conditions, changes in tax laws, adequacy of reserves, credit and other risks associated with the Arizona Utility Group and/or other third parties, significant changes in interest rates and fluctuations in foreign currency exchange rates.

Further, certain statements, findings and conclusions in this Report are based on Pace Global’s interpretations of various contracts. Interpretations of these contracts by legal counsel or a jurisdictional body could differ.

Proprietary & Confidential i

TABLE OF CONTENTS

1. Executive Summary ................................................................................................................... 5

1.1. EPA’s Interim Goals Resulting from Building Blocks are Unreasonable, Inequitable and Unachievable ............................................................................................................................. 5

1.2. There are Reasonable Alternatives Available to the EPA ......................................................... 8

1.3. The Costs of Compliance under the EPA Building Block Scenario are Very Significant .......... 9

2. Assessment Overview ............................................................................................................. 11

3. EPA’s Building Block Approach and Implications for Arizona ................................................. 12

3.1. Assessment of Reasonableness of Building Block Assumptions ............................................ 13

3.2. The CPP’s Proposed Goal Levels are Inequitable for Arizona ............................................... 17

3.3. Interim Goals ........................................................................................................................... 18

3.4. Implications of Building Blocks on Arizona’s Electric System ................................................. 18

3.4.1. Arizona Generation Mix and Installed Capacity under the CPP Building Block Scenario .................................................................................................................. 20

3.5. Risks to Electric Reliability ....................................................................................................... 20

3.5.1. Arizona Natural Gas Demand under the EPA Building Block Scenario .................. 22

3.5.2. Arizona’s Natural Gas Transportation Requirements under the EPA Building Block Scenario .................................................................................................................. 23

3.6. Risks to Natural Gas Supply Reliability ................................................................................... 26

4. Alternative Scenario Assessment ............................................................................................ 27

4.1. Generation Mix and Capacity Needs by Scenario................................................................... 29

4.2. Proposed Modifications to EPA Building Blocks to Address Interim Goal Issues ................... 30

5. Cost Implications of the Clean Power Plan ............................................................................. 31

5.1. Cost of New Generation Infrastructure .................................................................................... 31

5.2. Cost of Fuel and Purchased Power ......................................................................................... 31

5.3. Cost of Natural Gas ................................................................................................................. 32

Proprietary & Confidential ii

5.4. Other Cost Implications ........................................................................................................... 34

5.4.1. Cost of Stranded Assets ......................................................................................... 34

5.4.2. Cost of New Natural Gas Infrastructure .................................................................. 34

5.5. Cost of Changing Coal Plant Operational Behavior ................................................................ 34

6. Conclusions ............................................................................................................................. 36

6.1. Summary of Key Recommendations ....................................................................................... 36

Appendix A: Scenario Assumptions ............................................................................................................ 37

EPA Building Block Scenario Assumptions ....................................................................................... 37

Alternative Scenario Assumptions .................................................................................................... 43

Appendix B: Power Market Analysis Methodology ..................................................................................... 44

Power Market Modeling ..................................................................................................................... 44

Dynamic Build Capacity Expansion .................................................................................................. 45

Escalation Rate ................................................................................................................................. 46

Appendix C: Fuel Market Analysis Methodology ........................................................................................ 48

GPCM-Based Natural Gas Market Modeling .................................................................................... 48

Model Structure and Capabilities ......................................................................................... 48

Dynamic Build Capacity Expansion ..................................................................................... 49

Geography and Granularity .................................................................................................. 49

Natural Gas and Power Analysis Integration ....................................................................... 50

Proprietary & Confidential iii

EXHIBITS Exhibit 1: Recommended Adjustments to the CPP Building Blocks and Justification ........................ 8

Exhibit 2: Emission Rates by Scenario v. CPP Proposed and Adjusted Goals .................................. 9

Exhibit 3: Summary of Cost Impacts of the Clean Power Plan (2013$) ........................................... 10

Exhibit 4: Application of the Building Blocks to Arizona .................................................................... 12

Exhibit 5: Arizona’s Existing Coal and NGCC Units and Transmission Infrastructure ..................... 14

Exhibit 6: Arizona Affected Coal Unit Capacity Factors (EPA IMP Analysis of CPP) ....................... 15

Exhibit 7: Reduction of 2012 Emission Required by Final Clean Power Plan Goals (%) ................. 18

Exhibit 8: Building Block Scenario Assumptions............................................................................... 19

Exhibit 9: Arizona Generation and Installed Capacity (2015-2030) .................................................. 20

Exhibit 10: Arizona Reserve Margins with and without Incremental Natural Gas Builds .................... 21

Exhibit 11: Projected Annual Arizona Natural Gas Need in EPA Building Block Scenario ................ 22

Exhibit 12: Monthly Arizona Natural Gas Need 2015 v. Projected 2030 Building Block Scenario ..... 23

Exhibit 13: El Paso North Projected Monthly Pipeline Flow v. Pipeline Capacity ............................... 24

Exhibit 14: El Paso South Projected Monthly Pipeline Flow v. Pipeline Capacity .............................. 24

Exhibit 15: Transwestern Projected Monthly Pipeline Flow v. Pipeline Capacity ............................... 25

Exhibit 16: Summary of Alternative Scenario Modeled ...................................................................... 27

Exhibit 17: Affected Coal Assumptions by Scenario ........................................................................... 27

Exhibit 18: Arizona New Capacity by Technology by Scenario (MW) ................................................ 28

Exhibit 19: Total Arizona Generation Mix in 2030 by Scenario (MWh) ............................................... 29

Exhibit 20: Emission Rates by Scenario v. CPP Proposed and Adjusted Goals ................................ 30

Exhibit 21: Capital Costs for New Natural Gas Generation by Scenario (2013$M) ........................... 31

Exhibit 22: Fuel and Purchased Power Costs, EPA Building Block vs. Arizona Glide Path .............. 32

Exhibit 23: Projected National Natural Gas Power Sector Demand (EPA Building Block) ................ 33

Exhibit 24: Projected Henry Hub Natural Gas Pricing, EPA Building Block v. Reference .................. 33

Proprietary & Confidential iv

Exhibit 25: Coal-Fired Power Plant Cycling Cost Range, $000 per Cycle ......................................... 35

Exhibit 26: EPA Building Block Scenario Assumptions ...................................................................... 37

Exhibit 28: Natural Gas Price and Regional Basis (2013$/MMBtu) .................................................... 38

Exhibit 29: Average Arizona Delivered Coal Prices (2013$/MWh) ..................................................... 39

Exhibit 30: Arizona Load Forecast Before Efficiency (MW) and Efficiency Assumed (%).................. 40

Exhibit 31: Capital New Resource Technology Parameters for Market Expansion ............................ 41

Exhibit 32: New Units Additions .......................................................................................................... 41

Exhibit 33: Affected Coal Unit Assumptions by Scenario ................................................................... 43

Exhibit 34: Pace Global Market Analysis Methodology ...................................................................... 45

Exhibit 35: Dynamic Build Simulation Logic ........................................................................................ 46

Exhibit 36: Pace Global’s Annual Deflator Series ............................................................................... 47

Exhibit 37: GPCM Reported Natural Gas Market Points (Gas Hubs) ................................................. 50

Exhibit 38: Natural Gas Model Overview and Power Market Integration Scenatic ............................. 51

Proprietary & Confidential 5

1. Executive Summary

Several of the prominent utilities operating in Arizona including Tucson Electric Power Company (TEP), Arizona Public Service Company (APS), Salt River Project (SRP), Unisource Energy Services (UES), and Arizona Electric Power Cooperative (AEPCO), collectively referred to as the Arizona Utility Group or (AUG), retained Pace Global to perform an assessment of the impacts to the state that could result from the implementation of the EPA’s proposed Clean Power Plan (also referred to herein as the “CPP”) and to provide comments and recommendations to the EPA on its proposed rule.

Under the EPA’s proposed Clean Power Plan, Arizona would be required to reduce the carbon dioxide (CO2) intensity of its power generation fleet by approximately 47% by 2020 and approximately 52% by 2030 in order to meet its goals. Arizona undoubtedly faces one of the most aggressive reduction requirements of all states, driven by the application of the EPA’s proposed building block approach to determining state-level emission goals. Pace Global’s analysis of Arizona’s compliance implications finds that the rule is neither flexible nor achievable and that implementation of this plan without modifications would result in severe impacts to the reliability of electric supply in the state and excessive cost implications for Arizona customers. The CPP is projected to increase fuel and purchased power costs by 40% and generation capital expenditures by 30% when compared to more reasonable alternatives

Pace Global conducted the following analyses:

Analyzed the reasonableness of key assumptions in the building block approach for Arizona; Assessed the potential costs to Arizona customers associated with implementing the Clean Power

Plan as proposed by the EPA. Analyzed an alternative path for the state to reduce the carbon intensity of its generation on a realistic

and achievable timeframe.

Pace Global’s major conclusions are:

1. The interim targets resulting from EPA’s building blocks are unreasonable, are inequitable for Arizona, and cannot be achieved without major reliability concerns.

2. The infrastructure needs and costs associated with implementing the Clean Power Plan as it currently stands are very significant over a relatively brief period of time.

3. Alternative interim and final goals that fully consider the remaining useful life of existing plants would achieve significant carbon reductions without jeopardizing grid reliability and result in a lower cost to ratepayers.

Each conclusion is discussed in more detail below and in the remainder of this report.

1.1. EPA’s Interim Goals Resulting from Building Blocks are Unreasonable, Inequitable and Unachievable

Building block 1 cannot be technically achieved: EPA’s assumption of a six percent efficiency improvement for operating coal plants is highly speculative and technically impossible, especially noting that the rule, as proposed, would not account for efficiency improvements made to date.

Building block 2 should account for plant useful life and result in reasonable timeline for compliance: In the computation of the goal, the application of building block 2 accounts for 73% of Arizona’s total reduction requirement. Reductions from this building block would be required by 2020,


as the EPA assumes that increased utilization of existing natural gas combined cycle units and proportional reduction in coal generation is an immediate measure to reduce CO2 emissions from electric generation. In Arizona, this would result in the elimination of all affected coal capacity in the state without considering the significant remaining useful life that some of these units have. This assumption ignores many realities of how the electric power system operates and how long it takes to add required infrastructure, the fact that transmission in the state would need to be reconfigured (expanded) to operate gas units at the levels that would be required, and the fact that massive amounts of new gas capacity and pipeline infrastructure would have to be built to maintain operating reserve margins in Arizona. Pace Global’s analyses show that all coal plants would have to retire due to this building block by 2020 using realistic assumptions of plant operation.

– Pace Global recommends that the EPA consider the remaining useful life of existing plants as well as applying a phase-in of the re-dispatch assumed by building block 2 over the 2020 to 2030 time period rather than assuming that this re-dispatch could occur by 2020.

Building block 3 needs clarification: Although building block 3 accounts for a smaller portion of Arizona’s overall reduction requirement, renewable generation can be an important compliance strategy for Arizona in meeting its goals under the Clean Power Plan.

– Consistent with the treatment of renewables under virtually all state renewable energy standards, the EPA should clarify that renewable generation should be accounted for at the point of delivery and not the source of generation for use in compliance purposes under building block 3.

Building block 4 is not reasonably achievable: The application of building block 4 drives 15% of Arizona’s reduction requirements. This building block was apparently developed by considering what aggressive states have achieved over the past few years in energy efficiency reductions. There is no evidence that achieving levels of 1.5% per year can be maintained over a period of 10 or more years. The EPA’s approach selects far too aggressive efficiency levels for goal calculation and does not consider what Arizona has already achieved nor factor in the ability of states to meet the 1.5% annual reduction continuously for more than a decade.

– Pace Global recommends that the EPA adjust building block 4 to consider a 0.6% annual efficiency improvement rather than 1.5% when establishing overall target levels. This benchmark would be more in line with studies of achievable efficiency penetration levels.

The CPP is Inequitable for Arizona

The goals are particularly severe for Arizona and would require all affected coal in the state to be eliminated by 2020.

The state of Arizona must reduce its carbon levels by 47% by 2020 and 52% by 2030 from current levels, which is one of the highest reductions in the country, with all but 10 states having less than a 40% reduction by 2030. The fact that the vast majority of the reductions are required for the interim goal means that the plan must effectively be implemented by 2020, a virtually impossible task.

EPA’s Interim Targets Imply Major Reliability Concerns in Arizona

Pace Global’s analyses of the CPP Building Block application indicate that all of the non-tribal coal in the state would be retired by 2020. This would drive reserve margins in the state negative by 2020


without the addition of significant new capacity in the state. To maintain reserve margins, Pace Global estimates that by 2020, at least 2.4 GW of incremental natural gas generation capacity costing around $2 billion1 (over baseline expected needs to meet load growth and account for planned retirements) would be needed within a period of three years. This is virtually impossible to plan, permit and construct in Arizona.

The direct application of the four building blocks would require momentous changes to Arizona’s coal fleet by 2020. Approximately 3,316MW of coal fired generation would have to be retired by 2020, on top of already planned retirements and re-powerings, implying a stranded investment of over $3 billion (2013$) in 2020.

Most coal generation is in the eastern part of Arizona, while most of the existing natural gas generation is in western Arizona. Transmission capacity has been built out to serve load from the existing capacity sites. Without additional transmission infrastructure investment, which can take five to ten years to develop and construct, electric reliability and deliverability could be severely compromised.

There are two main natural gas pipelines serving Arizona, and one is already near capacity throughout the year, with both near capacity during peak periods. The CPP building blocks imply that a more than 3-fold increase in natural gas demand by the power sector alone would be expected by the early 2020s, driving the need for pipeline upgrades. Without additional pipeline infrastructure that can take four or more years to develop and construct, current pipeline capacity would be overwhelmed, and electric, as well as consumer natural gas, reliability and deliverability could be severely compromised.

1 This is generation cost only. Cost for additional transmission and gas pipeline infrastructure have not been specifically estimated in this analysis, but would be significant.


1.2. There are Reasonable Alternatives Available to the EPA

The significant cost, reliability and timing constraints that would impact Arizona’s ability to comply with the proposed goals of the Clean Power Plan can be mitigated by the application of several changes to the building blocks, which would result in a more gradual, but ultimately significant reduction in the CO2 intensity of generation. These specific recommendations and justification are summarized in Exhibit 1.

Exhibit 1: Recommended Adjustments to the CPP Building Blocks and Justification

Recommended Change for the EPA Rationale

Building Block 2

(1) Exclude coal plants from NGCC re-dispatch if they are 40 or less years old as of 2030.

(2) Evenly phase in the re-dispatch assumed by building block 2 over the 2020 to 2030 time period rather than assuming that this re-dispatch could occur by 2020

This would both account for the useful life of plants and result in a more feasible timeline to address infrastructure issues and costs (including stranded) associated with the required significant generation switching required.

Building Block 3 Clarify that all renewable generation should be accounted for at the point of delivery and not the source of generation.

This would enable states to rely on regional resources for compliance and is consistent with virtually all existing state renewable energy standard legislation.

Building Block 4 Adjust building block 4 to assume a 0.6% annual efficiency improvement rather than 1.5%.

This penetration level is more consistent with a reasonable achievable level for purposes of target setting and would allow states to rely on efficiency as a compliance mechanism, providing flexibility.

Source: Pace Global.

In applying these recommended changes to the EPA’s building block approach, Pace Global determined recommended goals for Arizona that would:

Reduce the carbon emission intensity of generation in Arizona by around 35% by 2030, with an adjusted final 2030 goal of 942 lbCO2/MWh versus the EPA’s goal of 702 lbCO2/MWh proposed;

Account for the useful life of coal plants; Provide adequate time to develop the transmission infrastructure to ensure grid reliability; Provide adequate time to develop the gas pipeline infrastructure to ensure gas supply reliability; Achieve these reductions at a much lower cost to the customer and avoid near-term rate shocks (as

depicted in Exhibit 3).

Exhibit 2 presents emission rates by year between now and 2030 for the EPA Building Block scenario and the Arizona Glide Path scenario based on Pace Global’s analysis. Both scenarios achieve significant reductions in carbon emissions. However, the Arizona Glide Path scenario offers a much more gradual path to meeting these reductions, without the cost and reliability concerns that would result from implementing the interim and 2030 targets for Arizona as currently proposed.


Exhibit 2: Emission Rates by Scenario v. CPP Proposed and Adjusted Goals


1.3. The Costs of Compliance under the EPA Building Block Scenario are Very Significant

Pace Global assessed the cost implications of the Clean Power Plan assuming the literal application of the building blocks as well as an alternative scenario to reducing emissions in a more reasonable and cost effective manner. Costs of the EPA building block analysis were compared to the alternative Arizona Glide Path scenario that accounts for the useful life of coal plants in the state to assess costs directly attributed to meeting Clean Power Plan compliance. This comparison is summarized in Exhibit 3 and includes the following key findings:

The fuel and purchased power component of costs for Arizona electric ratepayers are estimated to increase by 40% (with risk of higher impacts, depending on the impact of the plan on U.S. natural gas markets and pricing) under the EPA Building Block scenario versus the Arizona Glide Path scenario. This is due to fuel switching from lower-cost coal to higher-cost natural gas, as well as increases in the expected cost of natural gas over time as a result of substantially higher gas demand in the EPA Building Block scenario.

New capital expenditures associated with building gas plants are likely to be 30% higher in the Building Block scenario between 2020 and 2030 than the Arizona Glide Path scenario’s plan to phase coal out more gradually.

In addition, the EPA Building Block scenario would result in $3 billion in utility stranded costs in 2020, resulting in ratepayers paying twice for the same service.


Exhibit 3: Summary of Cost Impacts of the Clean Power Plan (2013$)

AZ Glide Path

Scenario

EPA Building Block Scenario

Delta (EPA BB – AZ Glide Path)

Percent Change

2020-2030 Average Fuel + PP Costs ($/MWh) $37.9/MWh $52.7/MWh $15/MWh 40%

2020-2030 Total Fuel + PP Costs ($Billion) $44.5B $62.2B $17.7B 40%

2020 – 2030 Gas Capacity (MW) 7,825MW 10,125MW 2,300MW 29%

2020-2030 Capital Cost Investment ($Billion) $6.2B $8.1B $1.9B 31%

Stranded Cost in 2020 Due to Early Coal Closures ($Billion)

n/a $3.04B n/a n/a

Note that the additional cost associated with new and upgraded electric transmission and natural gas pipeline infrastructure required

to meet Clean Power Plan goals are not included in this summary.


This analysis shows that customers would benefit greatly from a more moderate and gradual reduction in coal generation that accounts for the useful life of coal plants while still achieving significant reductions in carbon intensity.


2. Assessment Overview

Pace Global performed an assessment of the Clean Power Plan’s impacts on Arizona to analyze the following:

Analyzed the reasonableness of key assumptions in the building block approach for Arizona; Assessed the potential costs to Arizona customers associated with implementing the Clean

Power Plan as proposed by the EPA. Analyzed an alternative path for the state to reduce the carbon intensity of its generation on a

realistic and achievable timeframe. In performing the assessment, Pace Global reviewed the impacts of the Clean Power Plan on Arizona’s natural gas and electric power systems by assessing the implications of the prescribed draft rule and by comparing the plan with a more plausible alternative. Scenarios considered in this analysis include:

EPA Building Block scenario – literal application of EPA’s four building blocks resulting in 0 MW of affected coal capacity remaining in the state by 2020

Arizona Glide Path scenario – ~2,500MW remaining coal capacity in the state by 2030 Pace Global performed electric market dispatch analysis and fuel market analysis under these different scenarios to quantitatively assess the consumption, generation, cost, and infrastructure impacts specific to Arizona. To support the quantitative analysis, Pace Global deployed an hourly chronological dispatch model to simulate the economic dispatch of power plants within a competitive framework with the AuroraXMP platform. In its fuel market analysis, Pace Global utilized the Gas Pipeline Competition Model (“GPCM”) to conduct analysis of natural gas economics in North America. An overview of the modeling approach and assumptions used in the analysis are included as appendices to this report. The remainder of the report is organized into three major chapters as follows:

EPA’s Building Block Approach and Implications for Arizona Assessment of Alternative Scenario Cost Implications of the Clean Power Plan


3. EPA’s Building Block Approach and Implications for Arizona

The EPA defines the Best System of Emission Reduction (BSER) as four building blocks which are uniformly applied to actual 2012 baseline state generation profiles to determine individual state-level, rate based carbon dioxide goals (lbCO2 / MWh) under the Clean Power Plan. These building blocks, which the EPA characterizes as “reasonably achievable,” aim to account for individual states’ baseline generation mix. The resulting state goals, however, vary widely among different states in terms of the magnitude of reduction in emission rates required to comply and the expected cost and changes to generation mix that would be required to achieve them. As proposed, Arizona would be required to decrease its emission rate 52%, from 1,453 lbCO2/MWh in 2012 to 702 lbCO2/MWh by 2030, making it one of the most aggressive of all state goals in the proposed plan.

Exhibit 4 presents the calculation of the state’s goal by building block and the relative reduction applicable to each one.

Exhibit 4: Application of the Building Blocks to Arizona

Source: EPA.

• Arizona’s final goal: 702 lbs of CO2 / MWh by 2030

• Interim Goal: 735 lbs of CO2 / MWh average over 2020-2029

• Final Target Calculation:

• 2012 Baseline: 1,453 lbs CO2 / MWh

• Block 1 HR: 1,453 lbs CO2 / MWh 1,394 lbs CO2 / MWh 8% of total reduction

• Block 2 Disp.: 1,394 lbs CO2 / MWh 843 lbs CO2 / MWh 73% of total reduction

• Block 3 Ren.: 843 lbs CO2 / MWh 814 lbs CO2 / MWh 4% of total reduction

• Block 4 EE: 814 lbs CO2 / MWh 702 lbs CO2 / MWh 15% of total reduction

• Total Reductions:

• 1,453 lbs CO2 / MWh 702 lbs CO2 / MWh 52% reduction over baseline


3.1. Assessment of Reasonableness of Building Block Assumptions

Building block 1: Make Fossil Fuel Plants More Efficient – Building block 1 assumes that the average heat rate of affected fossil units2 decreases 6%. However, EPA’s basis for concluding that a 6% efficiency improvement for all affected units is possible, much less reasonably achievable, is not supported by the facts. The application of building block 1 would not account for efficiency gains made before 2012, and in effect would further disadvantage units that have already made efficiency improvements before this time. Interestingly, in the application of the building block approach under the proposed rule, Arizona would have to retire or otherwise eliminate all coal by 2020 to meet interim goals (pursuant to building block 2), which would essentially eliminate this building block as a compliance option for Arizona. Building block 2: Use Low-Emitting Power Sources More – Building block 2 assumes that affected existing and under construction NGCC units could increase dispatch up to a 70% capacity factor while proportionally backing down coal to reduce the emission intensity of the state’s generation mix. The application of this building block in determining state goals assumes that this occurs by 2020. Due to the large amount of NGCC capacity in Arizona, this building block has a significant impact on its state goal computation and would require significant and sudden shifts in the state’s generation mix to meet the resulting emission rate reductions. In fact, as applied in the EPA’s building block approach, the increased use of NGCC units in the state would displace all coal fired generation in the state, in effect forcing the retirement of all of the approximately 3,316 MW of coal capacity operating3 as of the 2012 baseline net already planned retirements and units slated for conversion to natural gas. There are several issues with the prescription of building block 2 for Arizona: (1) Existing NGCC Resources are Inadequate to Replace AZ Coal Retirements - Today, all Arizona utilities rely on the existing coal-fired and natural gas resources within in the state, including the existing natural gas merchant plants, to meet their summer peaking demand requirements. In addition to serving a portion of Arizona loads, Arizona’s merchant gas resources are also committed to meet summer peak demands in adjacent states. As a result, the early retirement of Arizona’s existing coal-fired resources by 2020 will necessitate the construction of new natural gas plants in order to maintain system reliability for 2020 and beyond. (2) Useful life of Fossil Generation is Not Considered – Arizona is home to some of the newest coal fired units in the country with the most recent units commencing operation as recently as 2009. These investments assume a long and useful life of 40 years or more. The application of building block 2 would require the retirement of virtually all if not all coal generation in the state by 2020, with no consideration to the useful life of the existing coal fleet. Retiring these units far earlier than a reasonable planned useful life would result in excessive stranded costs and ratepayers essentially paying twice for this generation capacity. (3) Transmission Infrastructure Would Not Support a Wholesale Change in the Generation Mix - The lack of transmission import capacity limits load serving entities from displacing all the retired coal units with the existing NGCC units, as shown in Exhibit 5. The majority of coal units and associated

2 Affected electric generating units are generally defined as currently operational or under construction in 2012, over 25MW and designed to operate more than one third of the time. 3 Arizona’s state goal only includes affected generating units on non-tribal land and therefore the Navajo generating station is not accounted for in the discussion of state goal computation and compliance implications.


transmission from which generation would shift is located in the far eastern and southern parts of the state while the existing NGCC units are located in and to the west of the Phoenix area. The transmission system works well for current operations, but the changes proposed by the EPA may well demand significant modifications. Exhibit 5: Arizona’s Existing Coal and NGCC Units and Transmission Infrastructure

Source: Pace Global, Ventyx.

(4) Increased Natural Gas Demand Would Strain Pipeline Infrastructure - Incremental natural gas demand resulting from the increased utilization of natural gas fired generation and the required new natural gas capacity needed to backfill the retired coal units to meet load would exceed the capacity of Arizona’s existing, but already heavily utilized natural gas pipeline network. Pace Global’s analysis indicates that compliance with Arizona’s interim goals would result in increased use of the El Paso system, as the alternate pipeline, Transwestern, is already 98% to 100% utilized. Both the El Paso and Transwestern systems would require expansion by the mid-2020s to maintain adequate supply capacity. Pipelines require a minimum of four years lead time from need determination to in service date. The timing of the CPP goals would require these expansion projects to begin soon to meet demand. Building block 2 accounts for over 70% of Arizona’s reduction required from baseline to its 2030 goal. The assumption that this generation switching could occur by 2020 is not feasible for Arizona due to in large part to the magnitude of investments in generation, transmission and pipeline infrastructure simultaneously, which require substantial commitments, permitting and construction lead times. Significant new infrastructure in the form of new electric transmission infrastructure, natural gas pipeline expansions, and generation infrastructure to maintain reliability would be needed. Further it would leave stranded investments in the state’s existing coal fleet that would impact electric rates.


(5) EPA IPM Modeling Found to be Inconsistent with Economic Utility Practices - The EPA’s analysis of the Clean Power Plan using the Integrated Planning Model (IPM) assesses several compliance scenarios. The Option 1 compliance for Arizona individually as a state is the most relevant comparison, considering that Arizona has no firm plans for regional compliance at this time. This modeling reflects 1,497 MW of coal-fired generation remaining online through 2030 and beyond while Arizona still meets interim and final proposed goals. The EPA modeling achieves this 2030 result through a decade of sub-optimal dispatch at the four units located at the Springerville Generating Station, the newest and generally most efficient of the coal-fired units currently operating in Arizona. As shown in Exhibit 6, the four existing units at the Springerville Generating Station maintain an annual capacity factor of 83% in 2016 and 2018. Starting in 2020, the output on Springerville Unit 3 drops to a 0% annual capacity factor with a maximum annual capacity factor of 26% in 2025 and 2030. Springerville Units 1 and 2 show annual capacity factors declining to 33% in 2020 and 36% in 2025 before returning to a 61% capacity factor in 2030. EPA’s modeling results in much smaller curtailments at the newest Springerville Unit, 4, with operation at an annual capacity factor of 66% in 2020 and a 72% annual capacity factor in 2025 and 2030. Pace Global finds the EPA analysis to be inconsistent with economic utility practice in its assumption that coal units, specifically Springervlle Units 1, 2 and 3, would operate for such an extended period of time at sub-optimal dispatch. This large reduction in the overall plant utilization at the Springerville Generating would result in an economic outcome that would favor shut down over operating the plant at average annual capacity factor of 40% over a 10 year period. Pace Global does not find this analysis to support Arizona’s ability to maintain any more than a very minimum capacity of the existing coal fleet online beyond 2019 while complying with EPA’s goals, particularly the interim goal. Exhibit 6: Arizona Affected Coal Unit Capacity Factors (EPA IMP Analysis of CPP)

Unit Capacity Factors

2016 2018 2020 2025 2030

Apache Unit 2 85%

Apache Unit 3 85%

Cholla Unit 1 84% 85%

Cholla Unit 2 83%

Cholla Unit 3 83%

Cholla Unit 4 84%

Coronado Unit 1 84% 84%

Coronado Unit 2 84% 84%

Navajo Unit 1-3* 83% 83% 83% 83% 83%

Springerville Unit 1 84% 84% 33% 36% 61%


Springerville Unit 3 82% 82% 26% 26%


*Note that Navajo units located on tribal land are not affected units under the proposed Clean Power Plan and therefore do not impact Arizona’s compliance.

Source: EPA Analysis of the Clean Power Plan, Option 1 – State


Finally, the modeling does not take into account region-specific technical issues. In particular, given the remote location of Arizona's existing coal units, it is anticipated that shutdown of these units may result in issues surrounding voltage and system stability. In the absence of a detailed transmission analysis, given Arizona’s demonstrated lack of flexibility, EPA would be unable to assess what units are necessary for the reliability of the electric grid.

Pace Global recommends that the EPA adjust the building block approach to exclude remaining useful life units from the re-dispatch calculation and phase in the application of building block 2 evenly over the 2020 to 2030 time period, similar to the application of building blocks 3 and 4. The additional timing will enable more cost effective decision making and reduce the cost and reliability impacts associated with stranded coal investments and simply defaulting to natural gas for compliance.

Building block 3: Use More Zero- and Low-emitting Power Sources – Building block 3 assumes that renewable generation meets a progressive state-assigned target based on regional build out trends by 2030 and that under construction nuclear units come online and at risk nuclear units stay online through 2030. The proposed rule is unclear as to whether or not renewable generation must be physically located in a state or delivered into a state, accounting for the fact that virtually all states that have renewable mandates implemented rely on out of state generation for compliance. Consistent with many legislated state renewable energy standards, the EPA should clarify that the renewable generation is accounted for at the final point of delivery and not the point of generation. Enforceability4 is one of the key criteria by which the EPA will assess state compliance under the Clean Power Plan. Noting this, it is very likely that most states will have to enact legislation of some type to align oversight agencies and generation owners and operators (and independent system operators where applicable) to ensure that responsible parties can be held to requirements under the state implementation plan. State renewable energy standards almost universally recognize delivered renewable energy from out of state sources for compliance. Legislation could only include the state RPS if the EPA were to accept that renewable energy delivered to a state from outside state borders could be recognized for compliance purposes under the Clean Power Plan. The nuclear portion of building block 3 defines certain new and “at risk” nuclear generation units that, should they not be online in the compliance timeframe, would very likely result in higher overall emission rates in select states, as the alternative generation would likely not all come from zero-emitting sources. The expected generation from these units was calculated to be on average approximately 6% of U.S. nuclear generation, which was uniformly applied to all states’ nuclear generation in developing proposed goals. This creates a significant inequity for states with nuclear generation that is not at risk. The EPA should amend this part of building block 3 to account for at risk units specifically and not apply a blanket average to all states that have nuclear generation. This would impact the application of building block 3 in determining state goals as well as the use of nuclear generation for compliance under the Clean Power Plan. The Palo Verde nuclear generating station in Arizona is not viewed as at risk, as it has several owners that rely on the plant to meet native load. However, through the application of building block 3, Arizona’s state goal is downwardly adjusted based on the national average of nuclear generation

4 The four general criteria by which the EPA proposes to evaluate SIPs are enforceability, achievement of state goals, verifiable emission reductions, and the process for regular reporting progress towards goal attainment.


deemed to be at risk. The EPA should clarify the ability of the state to document and count this unit for use in compliance towards its goals. Building block 4: Use Electricity More Efficiently – Building block 4 assumes that demand-side energy efficiency increases 1.5% annually from 2017 to 2030. This building block has the second largest impact on Arizona’s state goal derivation, driving a reduction of 15% of the rate reduction by 2030 versus the 2012 baseline. The EPA selected a 1.5% annual reduction based on a review of the top tier states’ efficiency performance in the 2012 baseline year, noting that three states actually achieved this level in one year and that nine other states had mandates to achieve this level for at least one year by 2020. Achieving a 1.5% efficiency improvement has been demonstrated in top tier states in a single year. However, the EPA’s assumption in building block 4 is that this high level is achieved by all states for the years 2017 to 2029. In the EPA’s Clean Power Plan technical support document on GHG Abatement Measures, only one source, the American Council for an Energy Efficient Economy (ACEEE), reported efficiency potential of 1.5% per year. The next highest efficiency potential was reported from the Electric Power Research Institute (EPRI), a non-partial source, in a 2014 study to be 0.6%, representing the high end of achievable efficiency penetration levels.5 Thus, even the EPA’s scientific support is not fairly supporting the fact that 1.5% efficiency penetration in all states is feasible, much less the achievement of this level consistently for more than a decade. Although a full legal analysis was not performed for this scope, Pace Global anticipates that this building block will be heavily scrutinized in the comment process and either reduced or eliminated from the building block approach in the final rule or challenged in the courts following the release of the final rule. For the purpose of goal setting, the EPA should lower the efficiency reduction level to 0.6% per year, in line with the EPRI efficiency potential study. Although Arizona does have strong efficiency standards adopted at the state level, the application of this building block just further lowers the state target and increases their already aggressive compliance burden. The EPA should ensure that goal setting does not become a compliance obligation, but rather enables entities to have flexibility to choose between renewable energy and energy efficiency to achieve compliance. Arizona’s existing efficiency program should be viewed as a compliance mechanism to manage its significant reduction requirements under the Clean Power Plan and particularly building block 2.

3.2. The CPP’s Proposed Goal Levels are Inequitable for Arizona

Implications for Arizona to comply with the proposed rule are disproportionately higher than nearly any other state. Arizona has one of the most severe reduction targets of all states covered under the Clean Power Plan, with a reduction below its 2012 baseline of 52%, while most states’ requirements are well below 50% with some less than 20% (Exhibit 7). Other states like Washington and South Carolina also have a large percentage reduction, although the drivers are different. Washington State has only a small amount of coal, so relatively minor generation switching measures can achieve compliance. South Carolina is impacted not by building block 2, as it has limited existing NGCC capacity, but rather by building block 3, due to the proposed new nuclear generation capacity in the state. Arizona stands out as the state most impacted by building block 2, due to its high coal generation and significant existing NGCC capacity.

5 EPA Technical Support Document “GHG Abatement Measures,” Table 5-7 “Summary of National EE Potential Studies”


Exhibit 7: Reduction of 2012 Emission Required by Final Clean Power Plan Goals (%)

Source: EPA, Pace Global.

3.3. Interim Goals

Interim state goals as proposed in the Clean Power Plan, which would need to be met on average over the 2020 to 2029 time period, are very close to the final goals and would require near-term generation changes including replacing most if not all of the coal capacity in the state, enhancements to the gas transportation infrastructure, and transmission infrastructure development. Arizona’s interim goal of 735 lbCO2 / MWh is 47% below its 2012 baseline emission rate. Because Arizona’s interim and final goals proposed are below the emission rate of even the most efficient natural gas generation, incremental renewable generation and energy efficiency over levels prescribed in the building blocks would be required to comply.

3.4. Implications of Building Blocks on Arizona’s Electric System

Pace Global assessed the implications of the CPP building block’s on Arizona’s electric system by quantitatively assessing compliance through an approximation of a literal application of the Clean Power Plan building blocks (generally following the EPA’s goal setting calculation). While this modeling approach shows Arizona complying with EPA’s interim and final goals, the results shown in this analysis are not achievable given the real world timeframes needed to construct new generation, transmission and gas pipeline infrastructure. The purpose of this Building Block scenario analysis is to highlight the potential reliability impacts for Arizona that are likely to result under the implementation of the Clean Power Plan. Exhibit 8 below summarizes the building blocks and modeling assumptions.


Exhibit 8: Building Block Scenario Assumptions

Building Blocks Modeling Assumptions Commentary on Modeling

1. Fossil plant efficiency improvements

Increase efficiency of existing coal plants by ~6%

Because of Arizona’s significant peaking natural gas capacity as of the CPP baseline in 2012, no coal operates in 2020 or beyond in order to meet building block 2

2. Coal-to-natural gas combined cycle (NGCC)

Increase utilization of all existing and new NGCCs up to 70% while proportionally reducing coal-fired generation

Existing combined cycle units are necessary to meet Arizona peak summer capacity requirements, not operate at a specific capacity factor value

3. Coal-to-low- or no-emitting sources

Increase renewables to Arizona state target of 4%, and assume no nuclear retirements

Renewable energy levels were assumed to meet EPA goal values; Palo Verde continues to operate

4. End-use energy efficiency Reduce demand-side energy use 1.5% annually through 2030

For Arizona this equates to ~12% by 2030

The study of the building block scenario focused on the generation supply and demand implications of the rule as well as the impacts to natural gas consumption and infrastructure utilization. An economic analysis of these impacts was performed as well to quantify costs associated with the implementation of the Clean Power Plan versus some of the alternative scenarios assessed. It is important to note that the study does not incorporate plant decommissioning expenses, specific transmission infrastructure or upgrade requirements, contractual take-or-pay provisions, or specific change in operating and maintenance costs at facilities.



3.4.1. Arizona Generation Mix and Installed Capacity under the CPP Building Block Scenario

Exhibit 9 shows the annual generation and installed capacity for Arizona from 2015 through 2030. Under the literal application of the Clean Power Plan building blocks, Arizona would be required to retire all of its existing coal-fired generation in 2020 to comply with EPA’s goals (except for tribal coal). The total elimination of coal-fired generation in Arizona required to meet the CPP interim targets would require an estimated 10 GW of new resource capacity from 2020 to 2030 to meet future reserve margins for Arizona’s peak summer season, with about 2.3GW directly attributed to CPP impacts and not load growth. The only coal-fired generation included beyond 2020 is sourced from the Navajo Generation Station that is on tribal land and not subject to Arizona interim and final goals. Exhibit 9: Arizona Generation and Installed Capacity (2015-2030)

Note that in-state generation dips slightly in 2020 as the new gas capacity including NGCC and peaking units does not completely replace baseload coal retirements and some economic imported purchased power is assumed to meet reserve margins in this period. Over time, new NGCCs do end up replacing most of the lost coal generation and the in-state generation rises.


3.5. Risks to Electric Reliability

The electric reliability issues from the Clean Power Plan are associated with both supply and deliverability. The elimination of existing coal-fired generation reduces overall electric supply. The assumption that incremental, lower-emitting generation resources can be permitted and brought online in a two to three year period from state plan development to compliance is not possible and would drive lower reserve margins. Even accounting for planned new builds between now and 2020, reserve margins would be right around 15%, meaning that the loss of most or all of Arizona’s coal fired generation would require virtually a MW for MW replacement of this generation to maintain safe reserve margins. An estimated 3.5 GW of new natural gas capacity by 2020 and over 10 GW by 2030 would be needed to meet reserve margin requirements in Arizona by 2030 following the retirement of all affected coal capacity within the state. This analysis assumes that Arizona utilities would not be able to rely on power imports from out of state to cover large shortfalls in generation capacity, meaning that new local natural gas builds would be needed in state to maintain system reliability. This assumption is based on the fact that the

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

Inst

alle

d C

ap

acit

y (M

Ws)

Nuclear Coal Existing Gas New Gas Hydro Wind Solar Other

-

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

Gen

erat

ion

(G

Wh

s)

Nuclear Coal Existing Gas New Gas Hydro Wind Solar Other


Arizona Corporation Commission’s (ACC) Integrated Resource Planning (IRP) rules state that utilities cannot rely on capacity that is not sourced from known generation for reserve margin requirements.6 Therefore, absent specific knowledge of generation resource and transmission infrastructure changes in neighboring states, out of state generation cannot be relied upon. Exhibit 10 shows the impact to Arizona’s reserve margins under compliance with state Clean Power Plan goals both with and without incremental new natural gas capacity. Exhibit 10: Arizona Reserve Margins with and without Incremental Natural Gas Builds


The assumption that coal generation can be one-for-one diverted to existing NGCC units is inaccurate, and the magnitude and timing which Arizona specifically would need to switch generation would make the state’s electric supply unreliable. The lead time for new transmission infrastructure is five to ten or more years. The recent North American Electric Reliability Corporation’s (NERC) report7 on the Clean Power Plan cites the need for a 10 to 15 year outlook for planning transmission development due to the time required for engineering, contracting, siting and permitting, as well as the various federal, state, provincial, and municipal approvals required. The CPP interim goals would allow for less than a 5 year outlook from state planning finalization until when the new transmission capacity would absolutely be needed. Since the CPP requirements will not be finalized until mid-2015 and state implementation plans will not be approved by EPA until mid-2017 or later, timing of the final state plan approval and the typical five-year

6 Arizona Corporation Commission Resource Planning and Procurement for 2011-2012, Docket No. E-00000A-11-0113, Decision No. 73884 7 Potential Reliability Impacts of EPA’s Proposed Clean Power Plan, Initial Reliability Review November 2014


timeframe to site and construct new power plants would result in real reserve margin declines, noting the sharp decrease in coal generation required under Pace Global’s analysis of the CPP Building Block scenario. Given these real world constraints, Arizona will need to seek relief from the CPP’s interim goals in order to maintain grid reliability and security. The EPA should include a reliability safety valve mechanism in the final rule. Even if the interim goals are delayed or state goals reduced, there is still a risk to reliability. Consistent with impacts, the EPA should include in its final rule circumstances under which compliance can be delayed to manage real time issues that will compromise electric reliability.

3.5.1. Arizona Natural Gas Demand under the EPA Building Block Scenario

Pace Global projects power sector natural gas demand in Arizona in the building block case to increase from 546 MMcf/d in 2015 to 2,088 MMcf/d by 2030, an almost four-fold increase from the power sector alone. This growth is driven by the increased utilization of existing NGCC units and incremental natural gas capacity additions to meet reserve margins. Growth in non-power sectors is expected as well, although as illustrated in Exhibit 11, these increases are dwarfed by the growth for power sector end use.

Exhibit 11: Projected Annual Arizona Natural Gas Need in EPA Building Block Scenario

Source: Pace Global

Arizona currently is a winter peaking market, but also exhibits a peak close to winter levels in the summer months as well. Although the increase in power sector consumption under the Building Block scenario is not anticipated to alter Arizona’s seasonal peaking profile, the peaks both in the winter and summer would increase dramatically as evidenced in the graphics in Exhibit 12.

0

500

1,000

1,500

2,000

2,500

3,000

201

5

201

6

201

7

201

8

201

9

202

0

202

1

202

2

202

3

202

4

202

5

202

6

202

7

202

8

202

9

203

0

MM

cf/d

Res Com Power Gen Ind LPP Fuel NGV


Exhibit 12: Monthly Arizona Natural Gas Need 2015 v. Projected 2030 Building Block Scenario

Source: Pace Global

3.5.2. Arizona’s Natural Gas Transportation Requirements under the EPA Building Block Scenario

Arizona benefits primarily, though not exclusively, from two long-haul pipelines, Transwestern and El Paso Natural Gas pipelines. Both of these pipelines are expected to remain highly utilized, as they currently are, to serve demand in New Mexico, Arizona and California and also to serve rising exports to Mexico. Given the current high capacity factor of pipeline usage in this region, the addition of Clean Power Plan induced natural gas demand for power needs in the southwest leads to concerns that the current pipeline system is inadequate to serve all demand during periods of peak usage. Pace Global’s pipeline flow model analysis indicates that under the Clean Power Plan, expansions would be needed to meet consumption needs and maintain reliability on both the northern and southern legs of the El Paso pipeline and on the Transwestern pipeline, as illustrated in Exhibit 13 through Exhibit 15. Pace Global’s pipeline flow model analysis indicates that under the Clean Power Plan, the following expansions would be needed to meet consumption needs and maintain reliability:

El Paso Southern leg would require expansion by the early 2020s if not sooner, as it is expected to exceed design flow by 2022 based on average flows.

El Paso Northern leg would require expansion by 2025. Transwestern pipeline, which is already over 95% utilized, would benefit from expansion,

although most of the incremental flows will impact the El Paso pipeline system.

0

500

1,000

1,500

2,000

2,500

3,000

3,500

Ja

n

Fe

b

Ma

r

Apr

Ma

y

Ju

n

Ju

l

Aug

Sep

Oct

No

v

De

c

MM

cf/d

2015: Monthly AZ Demand: CPP Building Block Case


0

500

1,000

1,500

2,000

2,500

3,000

3,500

Ja

n

Fe

b

Ma

r

Apr

Ma

y

Ju

n

Ju

l

Aug

Sep

Oct

No

v

De

c

MM

cf/d

2030: Monthly AZ Demand: CPP Building Block Case


2015 Monthly AZ Demand 2030 Monthly AZ Demand: EPA Building Block


Exhibit 13: El Paso North Projected Monthly Pipeline Flow v. Pipeline Capacity

Source: Pace Global, RBAC

Exhibit 14: El Paso South Projected Monthly Pipeline Flow v. Pipeline Capacity


0

500

1,000

1,500

2,000

2,500

3,00020

15

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

MM

cf/d

BAU Flow CPP Flow Capacity

0

500

1,000

1,500

2,000

2,500

3,000

201

5

201

6

201

7

201

8

201

9

202

0

202

1

202

2

202

3

202

4

202

5

202

6

202

7

202

8

202

9

203

0

MM

cf/d



Exhibit 15: Transwestern Projected Monthly Pipeline Flow v. Pipeline Capacity


Generators will need to secure existing pipeline capacity in order to ensure readily available supply. Even so, generators who are in the money and could generate, but who rely on interruptible pipeline capacity, may find themselves unable to dispatch during periods of high demand due to pipeline constraints. The pipeline constrained Northeastern U.S. experienced just such a situation during the 2013-2014 winter months. The gas-fired capacity expected to be built by 2020 will be contending with several large new sources of demand (e.g., LNG and pipeline exports, industrial projects), which will put upward price pressure on natural gas. This significant increase in natural gas demand coupled with inadequate natural gas storage and transportation infrastructure will ultimately lead to higher natural gas and power price volatility in the Desert Southwest.

1,170

1,180

1,190

1,200

1,210

1,220

1,23020

15

201

6

201

7

201

8

201

9

202

0

202

1

202

2

202

3

202

4

202

5

202

6

202

7

202

8

202

9

203

0

MM

cf/d



3.6. Risks to Natural Gas Supply Reliability

The dramatic projected increases in natural gas demand pose risks with regard to pipeline capacity able to deliver sufficient supply. In terms of whether there is sufficient time and incentive to build the pipeline that is needed, it is not likely. Pipeline is generally built on the basis of demonstrated firm demand resulting from an open season. Historically power generators have been hesitant to sign up for long term firm transportation capacity that is required to gain the needed investment in a timely manner. This results in a slower build-out of gas transmission, such as we see in the Northeast, where incremental pipeline capacity is not growing fast enough to mitigate periods of very high gas prices. As power generators’ share of gas consumption in Arizona and surrounding states grows, a similar situation to the Northeast region may develop in the Southwest, in which new pipeline capacity is hindered by hesitation for firm capacity commitments. Time to develop new pipeline infrastructure is at a minimum four years from the determination of need. The permitting stage can often significantly extend the process to five years or much longer. An extension of this timeline is very possible in Arizona and the greater Western U.S. as the amount of land owned by the military, federal government, state governments, and tribal nations is large. In Arizona, approximately 41% of land is owned by the federal government, almost 13% by the state, and about 27% belongs to tribal nations. All of this adds to the complexity of siting and constructing new capacity. Incremental expansions through compression upgrades can be realized in a two-year timeframe, but this only provides for small incremental capacity expansions. In fact, NERC8 specifically identifies Arizona as one state whereby the existing pipeline capacity is not adequate to handle incremental gas needs of the state under the CPP, consistent with Pace Global’s findings. Additionally, the natural gas supply situation in combination with literal application of the CPP building blocks appears to place Arizona in a precarious position. Given that Arizona is reliant on the supply of natural gas from 3 major pipelines, the prolonged disruption of service to one of these pipelines could prove devastating to Arizona residents in the absence of backup coal capacity.” Finally, grid reliability issues associated with increased renewable resources are not directly addressed as part of the EPA’s proposed building block approach. Based on recent industry studies9 and prior NERC reliability assessments, as the penetration of variable generation resources increases, maintaining system reliability will become more challenging. Given that Arizona would be required to retire all of its existing coal-fired generation in 2020 to comply with EPA’s goals, additional assessments, including interconnection-wide studies, will be needed as state implementation plans are developed to further understand potential reliability challenges that may indirectly result from the proposed CPP.

8 Potential Reliability Impacts of EPA’s Proposed Clean Power Plan, Initial Reliability Review November 2014

9 NERC-CAISO Joint Report: Maintaining Bulk Power System Reliability While Integrating Variable Energy Resources – CAISO Approach; other industry reports include those developed by the Integration of Variable Generation Task Force (IVGTF)


4. Alternative Scenario Assessment

Pace Global assessed the Arizona electric system under an alternative scenario that offers a more gradual reduction of coal generation in the state to reduce emissions. The purpose of this analysis was to identify the impacts to Arizona’s Clean Power Plan compliance and overall emission rate, while maintaining some of the most efficient coal generating stations in the state to reduce cost and reliability impacts.

Exhibit 16 presents a summary of the alternative scenario modeled.

Exhibit 16: Summary of Alternative Scenario Modeled

Scenario Remaining

Coal Capacity 2030 (MW)

Percentage of Coal Online in

2030 (%)* Rationale for Scenario

Arizona Glide Path scenario

2,542 77% Scenario maintaining useful life

of coal units

*Note, percentage based on the total affected coal capacity in Arizona excluding planned retirements and repowerings, including 3,316MW of the total 3,861MW operating today.

Source: AUG

A summary of the affected coal capacity and generation in the state is summarized in Exhibit 17.

Exhibit 17: Affected Coal Assumptions by Scenario

Capacity (MW)

Generation (MWh)

% Coal Capacity v. 2014

% Coal Generation v. 2014

Total Coal (2014) 3,861 24,801,925 100% 100%

Planned Retirements / Conversions by 2020 545 2,967,068 14% 12%

Planned Remaining Coal by 2020 3,316 21,834,857 86% 88%

EPA Building Block Scenario - Remaining Coal 2020 0 0 0% 0%

Arizona Glide Path scenario - Remaining Coal 2030 2,542 16,662,479 66% 67%

Note that only affected coal units in the state are included in these values.

Source: AUG and Pace Global


In comparison to EPA’s renewable target setting for Arizona under building block 3, the alternative scenario assumes significantly higher levels of installed renewable capacity from 2020 through 2030. Under this scenario, the assumed installed renewable capacity reflects a level more in line with the Arizona renewable energy standard which assumes 15% of a load serving entity retail load is sourced from renewable resources. Incremental renewable capacity additions assume approximately 85% solar and 15% wind. The higher levels of renewable generation, in combination with the reduction in retired coal fired capacity, reduce the amount of new natural gas fired generation needed to meet future load and reserve margins requirements in Arizona. Of course some new natural gas capacity would be expected to come online in Arizona, regardless of actions taken to reduce emissions, just to meet load growth. Exhibit 18 presents new capacity by technology for both scenarios. Exhibit 18: Arizona New Capacity by Technology by Scenario (MW)

Natural Gas Renewables

Total 2020

Cumulative Total 2020 -

2030

2020 Capacity Attributed to

CPP

Cumulative 2020-2030 Capacity

Attributed to CPP 2020

Cumulative 2020 - 2030

EPA Building Block

3,525 10,125 2,400 2,300 0 0

Arizona Glide Path scenario

1,125 7,825 0 0 1,000 3,462

Note: Arizona Glide Path scenario assumed as baseline case for comparison purposes to determine capacity additions attributed to the Clean Power Plan. Also, per the EPA’s building block approach, the renewable generation target would be met before 2020 and therefore no incremental additions are assumed over the 2020 to 2030 time period assuming the literal application of the building blocks.

Source: Pace Global


4.1. Generation Mix and Capacity Needs by Scenario

Exhibit 19 presents Arizona’s generation mix for both scenarios. To properly represent the generation mix of the state, the Navajo coal plant, which is not an affected unit for Arizona CPP compliance, is reflected as ~7% of the coal generation in the state that remains through 2030. Navajo represents the only coal remaining in the EPA Building Block scenario. The Arizona Glide Path scenario shows a declining relative share of coal generation through 2030, maintaining coal at 19% of generation by 2030. Exhibit 19: Total Arizona Generation Mix in 2030 by Scenario (MWh)

Source: Pace Global

Coal7%

Natural Gas65%

Hydro5%

Nuclear20%

Solar3%

Wind0%

EPA BB

Coal19%

Natural Gas51%

Hydro5%

Nuclear18%

Solar6%

Wind1%

AZ Glide Path


4.2. Proposed Modifications to EPA Building Blocks to Address Interim Goal Issues

Pace Global’s suggested modifications to the EPA’s building block approach include specific adjustments to building blocks 2 and 4 that would impact the level of Arizona’s goals. As referenced earlier in the report, these adjustments include:

– Pace Global recommends that the EPA consider the remaining useful life of existing plants as well as applying a phase-in of the re-dispatch assumed by building block 2 over the 2020 to 2030 time period rather than assuming that this re-dispatch could occur by 2020.

– Pace Global recommends that the EPA adjust building block 4 to consider a 0.6% annual efficiency improvement rather than 1.5% when establishing overall target levels. This benchmark would be more in line with studies of achievable efficiency penetration levels.

These building block adjustments would make compliance feasible under more extended timelines and would be more consistent with the emission trajectory of the Arizona Glide Path scenario. Exhibit 20 presents the annual emission rates for scenarios modeled and proposed and adjusted goals. The adjusted goals would result in a reduction of the interim goal for Arizona from the unachievable 735 lbCO2/MWh to 1,119 lbCO2/MWh. The final adjusted goal would be slightly higher than that proposed in the CPP at 942 lbCO2/MWh versus the proposed 702 lbCO2/MWh.

Exhibit 20: Emission Rates by Scenario v. CPP Proposed and Adjusted Goals



5. Cost Implications of the Clean Power Plan

There are significant cost impacts from new infrastructure investments, including power and natural gas infrastructure, operational changes to the generation fleet, and the recovery of stranded investments. Natural gas pricing would likely increase as well, noting the increased reliance on natural gas to meet power demand nationally. All costs are presented in this section in real 2013 dollars unless otherwise noted.

5.1. Cost of New Generation Infrastructure

The cost of new NGCC capacity is likely to be in the range of $1,000/kW, noting that installed costs will vary depending on the size of the project, technology selected, interconnection retirements, etc. Significant associated transmission investment could also be required depending on the location of the new capacity versus the load centers in the state. Capital cost assumptions are detailed in Appendix A. Pace Global estimates capital costs for new natural gas generation to meet CPP compliance, net of the expected required investment to meet load growth through 2030, to be $1.9 billion. This represents an approximately 31% increase in costs on average from 2020 to 2030 between the EPA Building Block scenario and Arizona Glide Path scenario, that would otherwise be borne by ratepayers under the Arizona Glide Path scenario gradual emission reduction plan.

Exhibit 21: Capital Costs for New Natural Gas Generation by Scenario (2013$M)

Total 2020 Cumulative Total

2020 - 2030

2020 Cost Attributed to

CPP

Cumulative 2020-2030

Cost Attributed to

CPP EPA Building Block

$3,000 $8,067 $1,991 $1,900

Arizona Glide Path

$1,009 $6,167 $0 $0


5.2. Cost of Fuel and Purchased Power

The EPA Building Block scenario would require a wholesale retirement of the coal capacity in Arizona, shifting the cost of fuel for ratepayers from coal to more expensive natural gas. Pace Global estimated costs of both fuel and purchased power for electric ratepayers in Arizona Glide Path scenario and the EPA Building Block scenario. Noting the uncertainty associated with the long-term impacts to natural gas pricing associated with the incremental demand under a Clean Power Plan compliance scenario, these costs were assessed both with and without the impacts of potential higher gas pricing (as described next). Exhibit 22 compares these costs estimated for the EPA Building Block scenario versus Arizona Glide Path scenario. This shows that the current building block plan would cost electric ratepayers up to 40% more in fuel and purchased power costs between 2020 and 2030 as a result of fuel switching and expected increases in natural gas prices over time.


Exhibit 22: Fuel and Purchased Power Costs, EPA Building Block vs. Arizona Glide Path


5.3. Cost of Natural Gas

Increases in natural gas demand resulting from implementation of the Clean Power Plan are likely to raise gas prices and overall consumer prices nationally and in Arizona. The impact of an additional gas demand required for compliance would come at a time when exports (liquefied natural gas and pipeline exports to Mexico) are also growing rapidly. Pace Global estimates that all other things constant, power sector consumption could increase from approximately 24 Bcf/d in 2015 to 47 Bcf/d by 2030 under the EPA Building Block scenario. This is shown in Exhibit 23.

$0

$10

$20

$30

$40

$50

$60

$70

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030

Fu

el a

nd

Pu

rch

ased

Po

wer

Co

st (

$/M

Wh

)

AZ Glide Path Building Block Building Block - High Gas


Exhibit 23: Projected National Natural Gas Power Sector Demand (EPA Building Block)


Pace Global estimates that this incremental increase in national consumption levels would result in price increases nationally at the Henry Hub pricing point. When compared to a reference case outlook of Henry Hub pricing, the price increase under the Clean Power Plan is around $1.35 on average between 2020 and 2030, with deltas in the $1.50-2.00/MMBtu range by the end of the 2020s. Henry Hub pricing projections for the reference case and the EPA Building Bock scenarios are presented in Exhibit 24.

Exhibit 24: Projected Henry Hub Natural Gas Pricing, EPA Building Block v. Reference


In addition, the market has recently seen a sharp increase in the volatility of market prices in regions where shale gas development has outstripped the infrastructure capability to deliver to markets where the gas is needed. Price spikes in the northeast last winter, for example, exceeded $123.50/MMBtu. As

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

50,000

201

5

201

6

201

7

201

8

201

9

202

0

202

1

202

2

202

3

202

4

202

5

202

6

202

7

202

8

202

9

203

0

MM

cf/d

$0.00

$1.00

$2.00

$3.00

$4.00

$5.00

$6.00

$7.00

$8.00

201

4

201

5

201

6

201

7

201

8

201

9

202

0

202

1

202

2

202

3

202

4

202

5

202

6

202

7

202

8

202

9

Rea

l 20

13$/

MM

Btu

EPA Building Block Case Reference Case


demand increases rapidly, this issue could become a much wider problem than in just the northeast. Volatility could be very significant with growth in power sector natural gas demand projected under the Clean Power Plan.

5.4. Other Cost Implications

5.4.1. Cost of Stranded Assets

The retirement of some or all coal-fired generating capacity in Arizona that would be necessary to meet the interim and final goals of the CPP would result in significant stranded asset value for utilities in Arizona. This stranded value is only exacerbated by recent investments in control technology and plant upgrades in recent years. Depending on the treatment, these costs would be recovered through higher rates for Arizona customers. The stranded asset value under the building block scenario is estimated to be $3.8 billion in 2020 ($3.0 billion in 2013$), resulting from the retirement of all coal generation in the state. Again, the ultimate treatment of these stranded costs could result in severe rate impacts for Arizona customers in or around 2020.

5.4.2. Cost of New Natural Gas Infrastructure

Generally speaking, new gas pipeline construction costs have averaged $155,000 per inch-mile, according to a 2013 U.S. pipeline economics study conducted by Oil and Gas Journal. For smaller pipes, less than 12 inches in diameter, costs are assumed to range from $20,000 to $70,000 per inch-mile. Adding capacity potentially can cost less if the design capacity of the pipeline allows for the addition of incremental compression. The northern leg of El Paso Natural Gas pipeline is likely to become highly constrained by 2020 and in need of expansion. At a minimum, the 330 mile length of El Paso in Northern Arizona, which varies between a 30 and 36 inch diameter and averages approximately 2.5 Bcf/d in throughput capacity, would need to be expanded by 500 MMcf/d no later than 2023 at a likely maximum cost of $335 million.10 If a 12 inch or smaller diameter pipeline can be used (which has a median $45k/mile cost, according to the same Oil and Gas Journal study), then at a minimum the additional infrastructure cost would be $97 million. The range for a gas pipeline infrastructure upgrade on the most affected portion of pipeline (El Paso Arizona North) is between $97 and $335 million dollars. This estimate is illustrative of one expansion required for major a major pipeline and not exhaustive of all other upgrades that would be required for both major pipelines and smaller distribution systems.

5.5. Cost of Changing Coal Plant Operational Behavior

In addition to coal plant retirements, compliance with the Clean Power Plan could include the operation of coal plants at less than economic dispatch. This could include startup/ shutdown cycles and load following cycles. For coal-fired units, cold starts are often defined as when a unit is offline over 40 hours. Warm starts are commonly defined to include starts occurring after the unit has been offline from five to forty hours. Hot starts are generally those occurring within five hours of a unit going offline. While each of these cycles causes some measure of wear and tear in excess of steady state operation, cold starts are generally the most damaging and load following the least damaging. As mentioned above, cycling any unit results in increase equipment wear and therefore cost. These impacts generally manifest themselves in some combination of the following measurable effects:

10 These figures are derived from the following: ($155,000 $/inch-mile) * (330 miles) * (6.5 inch-equivalent of additional pipeline capacity needed).


Decreased reliability evidenced by increased equivalent forced outage rate (EFOR) and

increased cost of replacement energy; Derating resulting from damage; Increased capital and maintenance costs to repair increasingly worn parts; Increased fuel consumption, either from increased startups and/or higher heat rates resulting

from part load operations; Other startup costs i.e. chemicals, auxiliary power, manpower; Shorter unit economic life.

Depending upon the cycling type, from 49-62% of the cycling cost results from increased capital and maintenance costs associated with increased maintenance frequency, inspections, and repairs. Moreover, 23-29% of the cost increase results from forced outages requiring not only repairs (parts and labor), but also purchase of replacement power. Arizona’s coal-fired units were designed for baseload service and therefore do not cycle well. As a result, load cycles requiring startups and shutdowns will mostly be met with combustion turbine-based power plants, either combined cycle or peaking units. A limited amount of load cycling while the unit is online can be feasible and has relatively low operating costs. As evidenced in Exhibit 25, not only are expected load cycling costs much lower than the startup / shutdown cycles, the range of expected costs, and therefore the certainty around those costs, is higher than load following operations. The amount of load cycling that is possible is, however, constrained by minimum load requirements, O&M cost impacts, and emissions issues associated with running coal plants at low loads.

Exhibit 25: Coal-Fired Power Plant Cycling Cost Range, $000 per Cycle

Note: Typical 500 MW conventional coal-fired power plant. Values in 2008$

Source: Coal Power Magazine, Intertek-Aptech, Pace Global.

0

50

100

150

200

250

180 MWmin load

1-23 hrsOffline

24-120 hrsoffline

> 120 hrsoffline

LoadFollow

Hot Start WarmStart

Cold Start


6. Conclusions

6.1. Summary of Key Recommendations

Arizona’s interim goals proposed in the Clean Power Plan, which require a shutdown of the entire, non-tribal coal fleet in Arizona by 2020, cannot be met. A state plan for compliance and associated regulation that likely will be likely be needed for enforcement will not be in place before June 2017, and three years is inadequate to plan, coordinate, permit, construct and operate the natural gas plants necessary to maintain statewide reserve margin requirements, the transmission infrastructure required to deliver the gas fired generation to customers and the natural gas pipeline capability to supply the generating facilities.

Arizona is one of the hardest hit states in the country by the Clean Power largely because of inappropriate assumptions about the ability to redispatch from coal to gas by 2020 that does not account for the major transmission and infrastructure required to do so. Further, the EPA’s energy efficiency assumptions used in goal setting are well above levels that have been maintained to date for any significant length of time. This creates a high bar for compliance rather than a reasonably achievable efficiency benchmark for goal determination. Even the EPA’s own analysis of Arizona’s ability to comply with the proposed goals has coal capacity running below operating minimums.

These inequities and impossible to meet targets can be best dealt with through a number of measures:

Amend building block 2 to set a balanced more gradual reduction target for coal fired generation between 2020 and 2030 and account for the useful life of coal units that will give time to develop the needed infrastructure to build needed gas generation in the state and reduce stranded cost impacts to ratepayers.

Amend Building Block 3 to ensure that renewable generation is counted at the point of delivery.

Recalculate Building Block 4 to include energy efficiency measures of 0.6% per year rather than 1.5% per year.

Reset the interim and 2030 targets for Arizona consistent with these measures.

Making these adjustments would still achieve dramatic reductions in Arizona’s carbon footprint by 2030, but would do so in a way that would not jeopardize reliability of power and natural gas supply in the state and would avoid the likelihood of severe customer rate shock by 2020.


Appendix A: Scenario Assumptions

Pace Global performed power and natural gas market modeling for the EPA Building Block scenario and the alternative Arizona Glide Path scenario. Assumptions underlying these analyses are presented in this section.

EPA Building Block Scenario Assumptions

The EPA Building Block Scenario assumes the following assumptions to literally model the EPA’s building blocks for Arizona.

Exhibit 26: EPA Building Block Scenario Assumptions

Building Blocks Modeling Assumptions Commentary

1. Fossil plant efficiency improvements

Increase efficiency of existing coal plants by ~6%.

Because of Arizona’s significant peaking natural gas capacity as of the CPP baseline in 2012, no coal operates in 2020 or beyond in order to meet building block 2.

2. Coal-to-natural gas combined cycle (NGCC)

Increase utilization of all existing and new NGCCs up to 70% while proportionally reducing coal-fired generation.

Existing combined cycle units are necessary to meet Arizona peak summer capacity requirements, not operate at a specific capacity factor value.

3. Coal-to-low- or no-emitting sources

Increase renewables to Arizona state target of 4%, and assume no nuclear retirements.

With under construction renewables, Arizona already exceeds 4%; therefore current levels of ~5% are assumed; Palo Verde continues to operate.

4. End-use energy efficiency Reduce demand-side energy use 1.5% annually through 2030.

For Arizona this equates to ~12% by 2030.

The following tables present key assumptions underlying the analysis.


Exhibit 27: Natural Gas Price and Regional Basis (2013$/MMBtu)

Year Henry Hub N. Arizona Basis S. Arizona Basis

2015 3.77 0.05 0.15

2016 3.88 0.02 0.14

2017 4.08 0.05 0.18

2018 4.33 0.06 0.23

2019 4.67 0.03 0.18

2020 5.39 0.11 0.30

2021 5.57 0.09 0.24

2022 5.63 0.12 0.25

2023 5.50 0.15 0.24

2024 5.49 0.16 0.23

2025 5.53 0.17 0.21

2026 5.55 0.27 0.30

2027 5.58 0.34 0.35

2028 5.64 0.38 0.38

2029 5.73 0.42 0.40

2030 5.77 0.57 0.53

2031 5.84 0.60 0.53

2032 5.93 0.65 0.57

2033 6.01 0.67 0.57

2034 6.05 0.68 0.57

2035 6.09 0.71 0.59

Source: Pace Global


Exhibit 28: Average Arizona Delivered Coal Prices (2013$/MWh)

Year Coal Price ($/MWh)

2015 23.2

2016 23.4

2017 23.7

2018 23.6

2019 23.6

2020 23.5

2021 23.3

2022 23.3

2023 23.2

2024 23.2

2025 22.7

2026 22.7

2027 22.6

2028 22.6

2029 22.5

2030 23.4

Note: Delivered coal prices to individual plants in Arizona range from approximately $21.5/MWh to $27.5/MWh.

Source: Pace Global


Exhibit 29: Arizona Load Forecast Before Efficiency (MW) and Efficiency Assumed (%)

Year Average Peak Building Block 4

Efficiency %

2015 10,469 20,055

2016 10,788 20,669

2017 11,106 21,283 1.5%

2018 11,425 21,896 1.5%

2019 11,743 22,509 1.5%

2020 12,062 23,121 1.5%

2021 12,380 23,734 1.5%

2022 12,698 24,346 1.5%

2023 13,017 24,960 1.5%

2024 13,336 25,574 1.5%

2025 13,654 26,187 1.5%

2026 13,973 26,800 1.5%

2027 14,298 27,427 1.5%

2028 14,631 28,069 1.5%

2029 14,972 28,726 1.5%

2030 15,321 29,398

2031 15,678 30,086

2032 16,043 30,790

2033 16,417 31,511

2034 16,800 32,248

2035 17,191 33,003

Note: Load forecast is gross economic demand and does not include any efficiency or demand side program assumptions. All scenarios assume that the building block 4 efficiency annual percentages are applied resulting in a cumulative efficiency savings of 11.4% by 2030.

Source: Pace Global and EPA


Exhibit 30: Capital New Resource Technology Parameters for Market Expansion

Technology

Early Capital Cost

Mid Capital Cost

Late Capital Cost

Early Levelized

Mid Levelized

Late Levelized

(2014-2016) (2017-2024) (2025-2030) (2014-2016) (2017-2024) (2025-2030)

$/kW $/kW $/kW $/kW-yr $/kW-yr $/kW-yr

CC (7FA) 1046 974 890 148 139 127

CT (FA) 735 685 624 115 109 100

Advanced CT (LMS 100) 1099 996 880 160 147 131

Solar PV* 2392 2012 1616 161 200 165

Wind 1.5 MW* 1909 1779 1632 183 212 197 *Wind and Solar Costs increase during the "mid levelized" period as tax benefits such as PTC is assumed to phase out and ITC reduces.

Source: Pace Global

Exhibit 31: New Units Additions

Owner Name Plant Name NERC Sub

Region Unit Status

Online Date

Primary Fuel

Prime Mover

Winter Capacity

(MW)

PacifiCorp Lake Side Power Plant BASIN Operating May-14 Gas CC 647.4

Kennecott Utah Copper Corp KUCC BASIN Under Const Aug-15 Gas GT 5.9

Gerlach Geothermal LLC San Emidio Project BASIN Under Const Jun-15 Geo GE 8.6

Durbin Creek Windfarm LLC Durbin Creek Windfarm BASIN Site Prep Sep-14 Wind WT 20.0

Willow Spring Windfarm LLC Willow Spring Windfarm BASIN Site Prep Sep-14 Wind WT 30.0

Hanford Peaker LLC GWF Hanford Combined Cycle CAL N Site Prep Sep-16 Gas CC 120.0

Henrietta Peaker LLC Henrietta Peaker*** CAL N Site Prep Sep-16 Gas CC 123.0 Contra Costa Generating Station LLC Oakley Generating Station ** CAL N Under Const Dec-16 Gas CC 624.0

Xeres Ventures LLC Santa Clara SC1 Data Center CAL N Operating Jun-14 LOil IC 9.0

Westlands Solar Farms LLC Westlands Solar Farm CAL N Operating Apr-14 Solar PV 23.0

Topaz Solar Farms LLC Topaz Solar Farm CAL N Under Const Jan-15 Solar PV 151.9

Topaz Solar Farms LLC Topaz Solar Farm CAL N Under Const Mar-15 Solar PV 92.0

Lax Arpt Central Utilities Plant LAX CAL S Under Const Dec-14 Gas CC 8.8

Lax Arpt Central Utilities Plant LAX CAL S Under Const Dec-14 Gas CT 6.6 Los Angeles Dept of Water & Power Scattergood CAL S Under Const Dec-15 Gas CC 309.0 Los Angeles Dept of Water & Power Scattergood CAL S Under Const Dec-15 Gas GT 190.0

Lax Arpt Central Utilities Plant LAX CAL S Under Const Dec-14 Other CA 2.2

Genesis Solar LLC Genesis Solar Energy Project CAL S Operating Mar-14 Solar SS 125.0

Desert Sunlight 300 LLC Desert Sunlight Solar CAL S Under Const Jun-14 Solar PV 25.2

Desert Sunlight 300 LLC Desert Sunlight Solar CAL S Under Const Jul-14 Solar PV 20.2

Desert Sunlight 300 LLC Desert Sunlight Solar CAL S Under Const Aug-14 Solar PV 18.9

Desert Sunlight 300 LLC Desert Sunlight Solar CAL S Under Const Oct-14 Solar PV 22.7

Desert Sunlight 250 LLC Desert Sunlight Solar CAL S Under Const Nov-14 Solar PV 25.2

SG2 Imperial Valley LLC Solar Gen 2 CAL S Under Const Dec-14 Solar PV 50.0

SG2 Imperial Valley LLC Solar Gen 2 CAL S Under Const Dec-14 Solar PV 100.0


Owner Name Plant Name NERC Sub

Region Unit Status

Online Date

Primary Fuel

Prime Mover

Winter Capacity

(MW)

Imperial Valley Solar 3 LLC Imperial Valley Solar CAL S Under Const Dec-14 Solar PV 400.0

AES Solar LLC Mount Signal Solar Farm CAL S Under Const Dec-14 Solar PV 109.0

Solar Star California XIX LLC Antelope Valley I Solar Project CAL S Under Const Oct-15 Solar PV 310.0

Solar Star California XX LLC Antelope Valley II Solar Project CAL S Under Const Oct-15 Solar PV 276.0

Rice Solar Energy LLC Rice Solar Energy Project CAL S Site Prep Jun-16 Solar SS 150.0 Los Angeles Dept of Water & Power Headworks Reservoir CAL S Under Const Dec-17 Water HY 4.0

Jawbone Wind Energy LLC Jawbone Wind Energy Project CAL S Site Prep Mar-15 Wind WT 39.0

Geolectric Power Co NM LLC Lightning Dock Geothermal DESERT

SW Site Prep May-15 Geo GE 6.0

Sexton Energy LLC Tangerine LFG Project DESERT

SW Under Const Jan-15 Renew IC 1.4

Copper Mountain Solar 2 LLC Copper Mountain Solar DESERT

SW Under Const Oct-14 Solar PV 30.0

Copper Mountain Solar 2 LLC Copper Mountain Solar DESERT

SW Under Const Dec-14 Solar PV 30.0

Tucson Electric Power Co H Wilson Sundt Generating Station

DESERT SW Under Const Dec-14 Solar PV 5.0

Sempra Generation Copper Mountain Solar DESERT

SW Under Const Mar-15 Solar PV 250.0

American Capital Energy Searchlight Solar DESERT

SW Under Const Jun-15 Solar PV 20.0

First Solar Inc Moapa Solar Project DESERT

SW Under Const Jun-15 Solar PV 150.0

First Solar Inc Moapa Solar Project DESERT

SW Under Const Dec-15 Solar PV 100.0

Torch Renewable Energy Red Horse 2 Wind DESERT

SW Under Const Jun-15 Wind/Sol

ar WT 70.0

Iberdrola Renewables Inc El Cabo Wind DESERT

SW Under Const Dec-15 Wind WT 298.0

Moapa Solar LLC Moapa Solar Energy Center DESERT

SW App Pending Sep-15 Solar PV 100

Moapa Solar LLC Moapa Solar Energy Center DESERT

SW App Pending Sep-16 Solar SS 100 Silver State Solar Power South LLC

Silver State South Solar Project

DESERT SW App Pending Dec-16 Solar PV 250

Arlington Valley Solar Energy I LLC

Arlington Valley Solar Energy Project

DESERT SW Proposed Dec-15 Solar SS 125

Arizona Public Service Ocotillo DESERT

SW Proposed Apr-18 Gas CT 525 Silver State Solar Power South LLC

Silver State South Solar Project

DESERT SW Proposed Dec-18 Solar PV 100

Pacific Hydro Inc Kingman Wind DESERT

SW Proposed Dec-17 Wind WT 10.2

Portland General Electric Co Carty Generating Station NWPP Under Const Jul-16 Gas CC 440.0

Dorena Hydro LLC Dorena Dam NWPP Under Const Oct-14 Water HY 5.2

Fairfield Wind LLC Fairfield Wind NWPP Operating May-14 Wind WT 10.0

Portland General Electric Co Lower Snake River Wind Energy Project NWPP Under Const Jun-15 Wind WT 267.0

Two Elk Generation Partners LP Two Elk Energy Park RMPA Site Prep Dec-16 Coal ST 290.0

Black Hills Corp Cheyenne Power Plant RMPA Under Const Oct-14 Gas CC 55.0

Cheyenne Light Fuel & Power Co Cheyenne Power Plant RMPA Under Const Oct-14 Gas CC 40.0

Cheyenne Light Fuel & Power Co Cheyenne Power Plant RMPA Under Const Oct-14 Gas CS 37.0

Public Service Co of Colorado Cherokee (CO) RMPA Under Const Sep-15 Gas CC 633.2

Haxtun Wind LLC Haxtun Wind Farm RMPA Site Prep Dec-14 Wind WT 28.8

Source: Pace Global and the AUG


Alternative Scenario Assumptions

Pace Global modeled an alternative scenario for reducing emissions statewide but also maintaining some coal generation in the state. The treatment of affected coal units in this scenarios is detailed Exhibit 32

Exhibit 32: Affected Coal Unit Assumptions by Scenario

Capacity (MW)

Generation (MWh)

% Coal Capacity v. 2014

% Coal Generation v. 2014

Total Coal (2014) 3,861 24,801,925 100% 100%

Planned Retirements / Conversions by 2020 545 2,967,068 14% 12%

Planned Remaining Coal by 2020 3,316 21,834,857 86% 88%

EPA Building Block Scenario - Remaining Coal 2020 0 0 0% 0%

Arizona Glide Path scenario - Remaining Coal 2030 2,542 16,662,479 66% 67%

Source: Pace Global and the AUG

The Arizona Glide Path scenario did not prescriptively model the EPA’s building blocks with the exception of building block 4 to ensure consistency in load across all analyses. The prominent assumptions presented in the tables above were also assumed for this scenario. No improvements in coal plant heat rates are assumed. NGCC units operated at economic dispatch levels. The renewable build out in Arizona was based on aggregate estimates by utility.


Appendix B: Power Market Analysis Methodology

Power Market Modeling

Pace Global deploys an hourly chronological dispatch model to simulate the economic dispatch of power plants within a competitive framework. Representations of hourly regional demand profiles and plant-level supply characteristics are included, as well as detailed assessments on the fundamental drivers of power plant dispatch within each relevant market area. Key components of our methodology include:

Load Forecast: Pace Global independently develops regional load forecasts (with stochastic uncertainty bands) based on the historic relationship between economic drivers, weather, and load.

Regional Fuel/Emission Projections: Pace Global develops independent projections of fuel and emission pricing inputs (with stochastic uncertainty bands) based on the fundamental drivers of each market and a comprehensive review of regulatory environments.

Renewable Generation Profiles: Pace Global analyzes the historic generation of renewable technologies throughout its modeling regions in order to characterize renewable generation profiles.

Bidding Function: Pace Global’s market simulations incorporate bidding behavior and scarcity premiums in our dispatch algorithm. Each region’s bidding function is based on hourly analyses of the historic relationship between prices and reserve margins

Dynamic Capacity Expansion: Gas-fired, wind, and solar capacity expansions are built dynamically when observed margins reach a specified threshold. – Creates boom/bust cycles that capture observed market behavior

A summary of the methodology with key inputs, algorithms, and outputs is shown in Exhibit 33.


Exhibit 33: Pace Global Market Analysis Methodology


Dynamic Build Capacity Expansion

Pace Global incorporates the dynamic simulation of additional economic capacity in our long term analyses. With this approach, incremental expansion is expected when economic conditions provide a sufficient rate of return for new units. Where net energy and capacity revenues together justify build of a new unit on the basis of a historic trend, a new unit is built. Sustained positive returns, generally stimulated by falling reserve margins and rising prices are expected to lead to capacity additions. The magnitude of the capacity expansion depends on the achieved Return on Investment (“ROI”) specific to the type of generating plant. Pace Global’s dynamic build logic is illustrated in Exhibit 34. This graphic illustrates how new capacity enters the market according to economic signals – these units are shown under the legend “Economic Expansion” (the units labeled “Additional Expansion” reflect announced units or units built on the basis of RPS or reliability requirements). For example, following an expected widening in system reserve margins over the period to 2009-2011, the system is expected to tighten during the 2011-2014 timeframe. In this example, we project that rising margins in the period 2011-2014 will send a signal causing a new plant to come online around the 2015 time frame. Following a temporary capacity glut, rising plant margins during the 2015-2018 period are unlikely enough to provide an unequivocal signal to new plant developers. In this case, a full build phase is not supported until the period from 2023-2026. From 2021, declining plant margins set in, reflecting the overbuild cycle. The dynamic expansion methodology is currently applied to incremental natural gas-fired combined cycles, natural gas-fired peakers, wind, and solar builds in the region, and is employed across all iterations of analysis. This allows all market simulations to incorporate the reactive behavior observed in the market to periods of sustained margins.

Aurora XMP

• Hourly Dispatch• Bidding• Dynamic Build• Detailed Market Representation

PlantGeneration

FuelPrices

Load

EmissionsCosts

Wind Generation

PlantVariable Costs

Plant Parameters

Inter-connections

PlantRevenues

Energy andCapacity Prices

Capital CostsFor New Entry

HydroGeneration

Stochastic Inputs


Exhibit 34: Dynamic Build Simulation Logic


Escalation Rate

Exhibit 35 shows Pace Global’s annual deflator series. Pace develops its market projections in real terms and converts prices to nominal values using the market rate implied by the yield on treasury bonds and similar maturity Treasury Inflation Protected Securities (“TIPS”). The yield quoted on treasury bonds is equal to the real yield plus inflation, while the yield quoted for TIPS is the real yield. Subtracting the yield of TIPS from the yield of Treasury bonds arrives at the market’s forward implied inflation rate.

0%

5%

10%

15%

20%

25%

30%

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

Res

erve

Mar

gin

0%

5%

10%

15%

20%

25%

30%

020406080

100120140160180

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

Res

erve

Mar

gin

Gro

ss M

argi

n ($

/kW

-yr)

0

100

200

300

400

500

600

700

020406080

100120140160180

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

New

Bui

lds

(MW

)

Gro

ss M

argi

n ($

/kW

-yr)

Build cycle

Falling Reserve Margin Increasing Energy and Capacity Margins for Plants

1 Decreasing reserve margins lead to… 2 Increasing gross margins for potential new entry…

3 Resulting in economic capacity builds

Reserve Margin Builds Gross Margin


Exhibit 35: Pace Global’s Annual Deflator Series

Year Deflator Series

2014 1.0164

2015 1.0331

2016 1.0500

2017 1.0672

2018 1.0863

2019 1.1058

2020 1.1256

2021 1.1494

2022 1.1738

2023 1.1987

2024 1.2260

2025 1.2540

2026 1.2826

2027 1.3118

2028 1.3417

2029 1.3723

2030 1.4036

2031 1.4356

2032 1.4683

2033 1.5018

2034 1.5356

2035 1.5701

Source: Pace Global and U.S. Treasury Department.


Appendix C: Fuel Market Analysis Methodology

GPCM-Based Natural Gas Market Modeling

In its fuel market analysis, Pace Global utilizes the Gas Pipeline Competition Model (GPCM) to conduct analysis of natural gas economics in North America. GPCM, developed and updated by RBAC, Inc., is a combination software-database system that allows Pace Global to quantitatively analyze the complex interactions among producers, pipelines, storage facilities, gas marketers, and consumers in the highly integrated North American natural gas industry. The primary output of GPCM is natural gas price forecasts and gas trading hub basis differentials to the Henry Hub, but has a range of other outputs including pipeline usage, transportation zone pricing customer receipts, storage balances, etc.

Model Structure and Capabilities

Mathematically, GPCM is a network model that can be diagrammed as a set of "nodes" and "arcs". Nodes represent production regions, pipeline zones, interconnects, storage facilities, delivery points, and customers or customer groups. The connections between these nodes are called arcs, which represent transactions and flows. Some of these are supplier deliveries to pipelines, transportation across zones and from one zone to another, transfers of gas by one pipeline to another, delivery of gas into storage, storage of gas from one period to another, withdrawal of gas from storage, and pipeline deliveries of gas to customers. GPCM dynamically solves for economic rents, allowing cheaper supplies to be used before more expensive supplies and enabling customers willing to pay more to be served before those willing to pay less. By including the entire system of North American gas production, transmission, storage, consumption, and imports/exports, GPCM optimizes gas flows in an economically sensible order to produce an economically efficient, market-clearing solution. GPCM contains more than 200 existing and proposed pipelines, 400 storage areas, 85 production areas, 15 liquefied natural gas (LNG) import/export terminals, and nearly 500 demand centers. GPCM can be adapted to model different scenarios based upon varying assumptions for projected gas supply and demand growth, among other variables. The model provides a “Base Case” scenario using existing pipeline tariffs, capacities, and normal weather for demand regions. This Base Case can be adapted to model the following factors:

Increases or decreases of projected demand by sector Increases or decreases of production capacity in traditional and unconventional areas Proposed pipeline projects or expansions Proposed LNG export terminals and capacity expansions New storage fields or increases in existing storage capacity

The output from GPCM consists of the following types of items, which can be exported to an Excel spreadsheet for further analysis and reporting:

Production and spot market prices by region Pipeline receipts from producers by zone Pipeline flows from zone to zone Transportation prices and discounting by pipeline and zone Transfers between pipelines at interconnects Injections into and withdrawals from storage Deliveries by pipelines to customers Gas supply available to each customer in each region Market clearing prices in each region


Dynamic Build Capacity Expansion

Pace Global has the capability to incorporate the dynamic simulation of additional pipeline capacity in our long term analyses. To the extent that the scenario under consideration requires that only “pre-programmed builds” come online (i.e., announced and under construction new pipeline or expansions that increase capacity), Pace Global has the ability to turn off the dynamic simulation switch. With this approach, incremental expansion is expected when economic conditions provide a sufficient rate of return for new pipeline capacity. Where utilization rates approach the capacity ceiling and there are economic opportunities to expand service, existing pipeline capacity can be expanded to increase revenue and reduce deadweight loss. Sustained positive returns are expected to lead to capacity additions. The magnitude of the capacity expansion depends on the rents that could be generated under an expansion scenario.

Geography and Granularity

GPCM covers the North American natural gas market, including Alaska, Canada, the continental United States, and Mexico. GPCM also contains a graphical display system to visually analyze interconnections, flows, and other output from the model. Demand forecasts can be manipulated by sector and by state. Supply sources can be manipulated by basin or play. Output data is provided on a monthly basis but can be aggregated up to annual averages. The forecasting horizon extends out to December 2035. Exhibit 36 below provides a list of natural gas market points reported out by GPCM (note that additional market points can be built into the model, as needed).


Exhibit 36: GPCM Reported Natural Gas Market Points (Gas Hubs)

Agua Dulce Hub Florida Gas, zone 3 SoCal Gas Algonquin, city-gates GTN, Kingsgate Southern Natural, La. Algonquin, receipts Henry Hub Southern Star, Tx.-Okla.-Kan. Alliance, into interstates Houston Ship Channel Stanfield, Ore. ANR, La. Iroquois, receipts TCPL Alberta, AECO-C ANR, ML 7 Iroquois, zone 2 Tennessee, La., 500 Leg ANR, Okla. Iroquois-Z1 Tennessee, La., 800 Leg Carthage Hub Katy Tennessee, zone 0 CEGT-South Kern River, delivered Tennessee, zone 6 delivered CEGT-West Kern River, Opal plant Texas Eastern, ELA CenterPoint, East Lebanon Hub-Ohio Texas Eastern, ETX Cheyenne Hub Leidy Hub Texas Eastern, M-1 (Kosi) Chicago city-gates Mich Con city-gate Texas Eastern, M-3 CIG, Rocky Mountains NGPL, Amarillo receipt Texas Eastern, STX Columbia Gas, Appalachia NGPL, La. Texas Eastern, WLA Columbia Gulf, La. NGPL, Midcontinent Texas Gas, zone 1 Columbia Gulf, mainline NGPL, STX Texas Gas, zone SL Consumers Energy city-gate NGPL, Texok zone TGP-Z1 100L Dawn, Ontario Niagara Transco, zone 1 Dominion, North Point Northern, demarc Transco, zone 2 Dominion, South Point Northern, Ventura Transco, zone 3 Dracut, Mass. Northwest, Can. bdr (Sumas) Transco, zone 4 El Paso, Bondad Northwest, s. of Green River Transco, zone 5 delivered El Paso, Permian Basin Northwest, Wyo. Pool Transco, zone 6 N.Y. El Paso, San Juan Basin Oneok, Okla. Transco, zone 6 non-N.Y. El Paso, South Mainline Panhandle, Tx.-Okla. Transwestern, Permian Basin Emerson, Viking GL PG&E city-gate Trunkline, ELA Florida city-gates PG&E, Malin Trunkline, WLA Florida Gas, zone 1 PG&E, south Waha Florida Gas, zone 2 Questar, Rocky Mountains Westcoast, station

Source: Pace Global

Additional information on GPCM can be found at www.rbac.com.

Natural Gas and Power Analysis Integration

Pace Global integrates its power and natural gas market analyses to account for the impacts of power sector consumption on natural gas infrastructure and pricing. Resulting natural gas demand from the power market analysis is run through GPCM to recalibrate pricing associated with the given consumption levels. This iterative process is performed until the resulting demand and pricing balance. Exhibit 37 presents the GPCM model parameters as well as the iterative process with the power market analysis used by Pace Global.


Exhibit 37: Natural Gas Model Overview and Power Market Integration Scenatic

Source: Pace Global

2014 1201 APS 111(d)Comments CO2Stnd-ExistingEGUs FINAL … · EPA Docket No. EPA-HQ-OAR-2013-0602...

Documents

Transcript of 2014 1201 APS 111(d)Comments CO2Stnd-ExistingEGUs FINAL … · EPA Docket No. EPA-HQ-OAR-2013-0602...