Chapter 7
Reconstructed Natural Gas-fired Stationary Combustion Turbines

The EPA notes that many of the comments in this chapter are similar to comments summarized in Chapter 7 of the response to comments for the January 2014 proposal. More detailed responses are provided in the RTC for the January 2014 proposal. The EPA also notes that many commenters combined comments for reconstructed and modified units. The responses included here are intended to only address reconstructed units.

Contents
7.1	Identification of BSER	2
7.1.1 NGCC Technology	2
Technical Feasibility	12
Costs	28
7.2	Determination of the Level of the Standard	29
7.2.1 Large Units Limit of 1,000 lb CO2/MWh; solicited comment on range of 950-1,100 lb CO2/MWh	32
7.2.2 Initial Performance Test in Addition to Ongoing Average	33
7.2.3 Load Following/Intermediate vs. Base Load Standard	37



Identification of BSER
Commenter 0215 stated that EPA did not explain why it believes that its January 2014 analysis, which focused exclusively on whether new Subpart KKKK units could meet its proposed standards, is relevant to the proposed standard's achievability for existing turbines that are reconstructed or modified. The commenter stated that the Combustion Turbine TSD limited its analysis to emissions data from 2007 to 2011 for units that commenced commercial operation after 2000 and completely ignored data from earlier years and older units.  
The EPA relied on the data for NGCC units that commenced operation in 2000 or after to determine what an achievable emission standard for reconstructed NGCC units. As discussed in Section IX.D.3.b of the preamble of this final rule, triggering the reconstruction provision requires a significant capital investment. It is reasonable to require a reconstructed combustion turbine to upgrade the turbine engine, the steam cycle, and integrate the two cycles to maximize efficiency. Many upgrade options are available such that reconstructed combustion turbines can meet the same standard as new combustion turbines. The standard is based on the performance of existing NGCC units and accounts for a wide variety of turbine sizes, operating conditions, site-specific conditions, and future degradation.
7.1.1 NGCC Technology
Commenter 0212 supported EPA's finding that modern, efficient NGCC technology is the BSER for stationary combustion turbines. The commenter noted that EPA solicited comment on whether the sales exemption in the proposed rule would be sufficient to account for the potential increased use of simple cycle combustion turbines due to the expected increased percentage of electricity generated from renewable generation in the future. The commenter believes that the threshold is sufficient. Increasing the exemption for lower efficiency units would undermine EPA's BSER determination. 
In the whitepaper, "Meeting the Needs of a Low-Carbon Power Grid: A study of dispatchable generation to partner with renewables," that was attached to their comment, the commenter discussed the role of simple cycle and combined cycle technologies in accommodating the supply fluctuations from increased renewables based on two recent studies. They also provided a supplemental analysis of the ERCOT power market. While simple cycle plants provide advantages for non-spinning reserve, and are easier to start-up and shutdown, these capabilities are not broadly needed to accommodate increased renewables. The reviewed studies show that combined cycle plants are projected to be dispatched for longer periods of time than simple cycle plants, and to stay online throughout the day, meeting ramping demands through load changes as opposed to multiple starts. The ERCOT data show that combined cycle power plants ramp up and down to satisfy fluctuating demand; while simple cycle units are used to meet infrequent peaks in demand. This will result in lower net generating costs and lower CO2 emissions than starting and stopping simple cycle plants multiple times throughout the day. 
The EPA acknowledges the commenter's support. The EPA notes that for this final rule, we erred on the side of caution and finalized a conservative and flexible set of requirements for all combustion turbines, including simple cycle and NGCC, to support the continued growth in renewable generation.
Multiple commenters (0146, 0150, 0187, 0199, 0203, 0275) noted that EPA had proposed "NGCC technology" as the BSER for modified and reconstructed natural gas combustion turbines, but was vague as to what "NGCC technology" includes. Commenter 0275 stated that the rulemaking docket does not clearly define "NGCC technology," and without a definition of "NGCC technology," an owner, operator or regulatory agency would not know precisely what equipment would need to be upgraded if a unit has been determined to have been modified or reconstructed. Commenter 0146 stated that precisely what "NGCC technology" may be has implications for evaluating the cost of technology and standard-setting. The commenters stated that EPA must clarify what it means by "NGCC technology." Commenter 0150 stated that NGCC technology varies greatly based on the age and size of the unit, so to give an across the board BSER like "NGCC technology" is arbitrary and capricious. Commenter 0203 stated that the term was too imprecise to constitute a readily apparent delineation of the BSER for this source category.
Commenter 0146 argued that EPA had failed to make any attempt to distinguish among turbines that were designed and constructed over a period of decades; consequently, the commenter objected to the Agency's determination that it is BSER without further elaboration or subcategorization. 
Commenters 0150 and 0203 stated that NGCC technology has evolved over decades and continues to improve. Commenter 0150 argued that it is not possible for every NGCC unit to achieve the emission rates of NGCC units constructed today, regardless of the work performed. Commenter 0199 stated that the EPA's "judgment is to be weighted against its statutory function and limitations, the record searched to determine if indeed its decisions and reasons therefore are themselves reasoned." See National Lime Ass'n, 627 F.2d 416, 431 (D.C. Cir. 1980), citing Essex Chemical, 486 F.2d at 434.
The commenter is mistaken about the purpose of the BSER determination. The BSER determination is use by the EPA to determine an appropriate technology to use as a basis for the corresponding emission standard. This final rule does not require the use of any particular technology. As discussed in Section IX.D.3.b of the preamble of this final rule, triggering the reconstruction provision requires a significant capital investment. It is reasonable to require a reconstructed combustion turbine to upgrade the turbine engine, the steam cycle, and integrate the two cycles to maximize efficiency. Many upgrade options are available such that reconstructed combustion turbines can meet the same standard as new combustion turbines. The standard is based on the performance of existing NGCC units and accounts for a wide variety of turbine sizes, operating conditions, site-specific conditions, and future degradation.
 Commenter 0211 argued that natural gas would be central to whether the series of proposed NSPS would be achievable. In its proposed rule for new stationary sources, EPA defined a NGCC unit to be BSER for subpart KKKK sources. The commenter stated that EPA cited three reasons for its determination: (1) the low carbon dioxide (CO2) emission rate associated with NGCC generation; (2) the use of a heat recovery steam generator in combination with a steam turbine to increase system efficiency, and (3) the cost-effective operation of such facilities. The commenter stated that in the proposed rule for existing stationary sources, EPA's analyses highlight the environmental attributes of NGCC. The commenter stated that EPA determined that NGCC units can produce as much as 46 percent more electricity from a given quantity of natural gas than steam EGUs and that generation from such units has the benefit of the inherently low-carbon content of natural gas. The commenter stated that in the current rule, EPA proposed to finalize the same standards for modified and reconstructed natural gas-fired stationary combustion turbines as the Agency did for new sources, noting that NGCC technology is one of the lowest cost forms of baseload and intermediate load electricity generation. While the commenter has not endorsed EPA's overall approach to the establishment of GHG standards for new units under CAA section 111 (Specifically, ANGA objects to the inclusion of carbon capture and sequestration (CCS) technology in the determination of BSER for any EGU) the commenter stated that natural gas-fired generation would be a readily-available, cost-effective technology to reduce CO2 emissions from the electric generation fleet while maintaining reliability. The commenter said that such generation will be a central part of efforts to comply with any final standard promulgated by EPA on the basis of this proposed rule or the proposed rules for new and existing stationary sources. The commenter stated that EPA must avoid creating barriers to use of natural gas or imposing measures that could unnecessarily drive up compliance costs. The commenter explained the benefits of natural gas to the environment and grid stability in comments filed on EPA's proposed rule for new stationary sources and specifically referenced and incorporated by reference those comments.  
The EPA acknowledges the commenter's support.
Multiple commenters (0224, 0227, 0257) requested that EPA exempt simple cycle CTs from the proposed rule. The commenters noted that EPA proposed that modified and reconstructed CTs be subject to the same emission standards as applicable to new units, "which are based on BSER being efficient NGCC technology." Commenter 0224 argued that it is unrealistic to force modified or reconstructed units to comply with the same standards as those for new units given the technical constraints existing units face in comparison to new units. Commenter 0224 stated that adopting the same NGCC standard would require that a CT installing an add-on control, such as SCR to lower NOx emissions, would be effectively forced to convert the CT into an NGCC. The commenter stated that the rule cannot be set to favor certain technologies and that the appropriate response would be to set higher limits with the range for modified or reconstructed CTs as compared to the limits based on a new NGCC in recognition of the differences between the two. 
Commenter 0224 suggested that EPA consider adopting the CT exemption initially proposed for new units. Commenter 0224 stated that, as noted there, simple cycle CT units that supply power to the grid are primarily used as peak units, and only infrequently as base or intermediate load power. Commenter 0224 stated that because simple cycle CT units serve as peaking units and for other system reliability purposes, but not as base load units, these limited and different functions distinguish them from NGCC units and other EGUs, which are designed to and regularly serve base and intermediate load functions. Commenter 0224 stated that exempting CTs from the proposed rules would not have a dramatic impact on emissions levels, even if the CTs, when used for peak or reliability purposes, temporarily exceed the proposed standards.
Commenter 0227 stated that, while it may be theoretically possible to add a heat recovery steam generator to any simple cycle turbine, many such units perform specific duty that does not justify such expense. Commenter 0227 stated that EPA proposed to deal with this through the applicability criteria, which limit the applicability of the standards to units whose production equals one-third of their potential output. Commenter 0227 cited their comments on the NSPS for new units recommending a simple exclusion for simple cycle units from these standards. Commenter 0227 stated that simple cycle machines perform valuable functions, and their wholesale replacement with combined cycle capacity simply because it is more efficient ignores the purposes for which such units were originally constructed. The commenter stated that EPA can and should make reasonable distinctions between sizes and types of facilities, and clearly propose to exclude simple cycle turbines from these requirements.
The non-base load subcategory provides adequate flexibility for simple cycle turbines that reconstruct. When an owner or operator is contemplating reconstruction, they will determine what percentage electric sales they anticipate in the future. Similar to newly constructed units, if they anticipate selling electricity in excess of the percentage electric sales threshold, NGCC (or an equivalent technology) will be required to comply with the base load subcategory emission standard.
Commenter 0245 stated that NGCCs need to have the flexibility of operating in simple cycle mode where conditions warrant and that those circumstances must be excluded from compliance determination calculations. The commenter stated that NGCCs are sometimes built in stages, with the heat recovery steam generator and steam turbine added after construction and installation of the combustion turbine. The commenter stated that they can also be designed to operate in either mode to provide market response flexibility and to provide for continued operations during repair or maintenance of the steam turbine.
 The EPA disagrees with the commenter that simple cycle operation in an NGCC unit should not be counted towards compliance demonstration calculations. The final rule does not set separate standards for simple cycle versus combined cycle units. The emission limit required for a unit is based on the percentage of the potential electric output sold to the grid. Further, the base load emission standard includes sufficient flexibility for limited operation of the combustion turbine engine without generation from the steam cycle.
Commenter 0245 argued that EPA had not considered the effect of the expected increase in renewables deployment and the resulting effects of increasing cycling of combined cycle units to support the integration of variable renewable resources. The commenter stated that thirty states have mandatory renewable energy standards with long term and near term incremental targets, and that seven states have voluntary programs. The commenter stated that new NGCCs will be built in an era that will require increased cycling due to the increased intermittency of non-baseload energy sources. The commenter stated that combined cycles operate most efficiently and effectively in a steady state while simple cycle units produce the lowest emissions during startup and shutdown, and that combined cycle technology is not BSER during frequent start and low capacity-factor operation.
In the final rule, NGCC has not been identified as the BSER for non-base load unit. NGCC has been identified as the BSER for base load units. As explained in Section IX.C.4.b, the BSER for non-base load units is the use of clean fuels. Therefore, NGCC units that cycle, have frequent starts and stops, and operate at part load conditions will only have to comply with an input-based standard based on the use of clean fuels as long as they remain below the percentage electric sales threshold. Our analysis of emissions data demonstrates that the base load emission rate of 1,000 lb CO2/MWh-g is readily achievable by NGCC units operating at ase load. See Section IX.C.4.b of the final rule preamble for more details on the EPA's rationale.  
Multiple commenters (0149, 0157, 0193, 0226, and 0250) argued that EPA has not explained how NGCC units, as BSER for modified and reconstructed natural gas turbines, can provide the same services as simple-cycle CTs. Commenters 0149 and 0226 stated that simple-cycle CTs are designed and built to provide specific grid support services, and argued that it is not appropriate to assume that NGCC units can set the standards for simple-cycle CTs or always be used to replace simple-cycle CTs. 
Commenter 0193 stated that, while NGCC units are also capable of rapid start and deployment, they are not nearly as agile as simple-cycle units and cannot fulfill the same role. Commenter 0193 stated that because of the added complexity of the associated steam cycle, a combined-cycle plant is unable to respond as quickly and flexibly as a simple-cycle turbine. Commenter 0193 stated that while many simple-cycle units are capable of multiple start and stop cycles per day, frequent start and stop cycles of NGCC facilities also induce significant thermal stress and associated operational cost and reliability tradeoffs that must be considered in the plant operation. As an example, commenter 0193 noted that an LMS100 simple-cycle gas turbine can achieve base load power of 100 MWs within 10 minutes of being called upon by the grid operators to fill an energy demand. Commenter 0193 stated that this same turbine could be loaded to approximately 25 MW (25 percent load) and any point in between while rapidly ramping between these load points, and that this fast start and ramping can be done multiple times a day to satisfy fluctuating energy demand. Commenter 0193 stated that by comparison, a 7F combined-cycle plant in a 2 on 1 configuration, in which the 2 gas turbines supply steam to a single steam turbine --a very common configuration in the United States--can also start in 10 minutes and achieve approximately 80 percent of gas turbine load for up to 200 MWs in 10 minutes. Commenter 0193 stated that the second gas turbine can be started along with the first turbine, or soon thereafter. The commenter stated that while starting the gas turbines, the exhaust heat is passing through the HRSG and begins generating steam, and that this process requires additional warm-up time. Commenter 0193 stated that the steam turbine would follow in load as the steam becomes available and can reach full load in roughly 30 minutes for an overnight plant shutdown, or take as long as 3 hours if the plant has been shut down for a few days and cooled to ambient temperature. Commenter 0193 stated that this illustrates how starting a combined-cycle plant is much more complex than starting a simple-cycle gas turbine and why it would not be done to satisfy an anticipated shorter term load demand.
Along with the ability for fast and frequent starting, commenter 0193 stated that the size of the plant also is a factor in dispatch and optimizing response to operational demand. Commenter 0193 stated that if a 75 MW load demand is anticipated, an LMS100 would be the most logical technology to satisfy that demand, and it would make no sense to start a plant capable of over 600 MWs to satisfy a relatively small transient anticipated demand. Commenter 0250 concurred, stating that NGCC plants are too large to meet their utility's needs for flexibility and redundancy. Commenter 0250 stated that even if operated without heat recovery, NGCC plants still have capacities starting at 225 MW for each turbine, far larger than the 100 MW capacity increments identified by their utility's system planning process. Commenter 0250 stated that the over-design problem worsens with consideration of the full capacity of these plants when operated in combined-cycle as they are designed (with heat recovery steam generation), which increases their capacity to a range of no lower than 305 MW. Commenter 0250 stated that if such a plant were to fail during a period of peak demand, they would not have the redundancy needed to meet that demand. Commenter 0250 stated that having multiple units with the proposed design configuration (i.e., four turbines at 100 MW capacity each), they can accommodate malfunctions or failures in their system (including failure of a unit at MPS), even during periods of peak demand, and that he incremental capacity growth of roughly an additional 100 MW per year also matches the load and resource needs resulting from expected growth and retirements in their utility system, while also spreading out capital requirements.
By building a plant with four GE LMS100 turbines, each with a capacity of 100 MW, commenter 0250 said their utility can deploy power over the range of loads needed to meet the project objectives while still meeting emission limitations (from 50-400 MW). Commenter 0250 stated that NGCCs cannot simultaneously provide 50 MW increments of power while also meeting emission limitations: At 50% load, the minimum capacity of NGCCs range from 112.5-169.5 MW, which is significantly greater than the lower range of power output needed for this project. Commenter 0250 stated that in order to provide the lower increments needed by their utility (50 MW), these NGCCs would need to be operated outside of their design range, which means that they would not meet emission limitations and could cause damage. Commenter 0250 stated that if they were to operate a unit above the level dictated by demand (contrary to prudent and accepted practice by a regulated utility) so as to run within emission limits that would still lead to wasted fuel and higher emissions (criteria pollutants and GHG). Commenter 0250 stated that even when over-producing power at 50% load, NGCCs would be at higher heat rate (lower efficiency) than the LMS100s at full load, thus the presumed low heat rates of a NGCC that form the basis for the pending proposed rules would not be obtained. Commenter 0250 stated that the GHG emissions (lb/MWh) and the heat rate (Btu/kWh) will be very high when NGCCs are operated at 50% capacity for longer periods. Commenter 0250 stated that simple-cycle LMS100s can be operated very efficiently at low loads (i.e., 50 MW) without damaging the equipment and with increased efficiency and lower emissions relative to the NGCCs. 
Commenter 0193 stated that the increased flexibility of simple-cycle units in comparison with NGCC units is also reflected in the operation of the power market. Commenter 0193 stated that NGCC units typically bid in the previous day for generation on the following day, while simple-cycle units can bid for generation on the same day--a mere hour ahead of the demand, or even go online with full output in less than 10 minutes in response to real-time market demands. Commenter 0193 stated that it would be prohibitively expensive for traditional NGCC units to bid on the same day. Commenter 0193 stated that while NGCC plants are able to respond to grid variability demands, simple-cycle units provide another layer of flexibility for rapid starts, stops, and grid fluctuations on a smaller scale than NGCC units can provide, and because these starts and stops correlate with lower efficiency and increased CO2 emissions, simple-cycle units used in this manner will not be able to meet the proposed standards.
Commenter 0157 stated that, with regard to reconstructed simple-cycle gas turbines, adding an HRSG is a question of space, design, and cost, given that simple-cycle units are designed to operate at low capacity factors. Commenter 0157 stated that installing an HRSG on a low-capacity factor unit, and one that may frequently operate on high-cost fuel oil, makes little economic sense. Commenter 0157 stated that EPA is not permitted to redefine the source. Commenter 0157 stated that simple-cycle and combined-cycle gas turbines are designed to serve different purposes, and that requiring the redesign of a reconstructed simple-cycle CT as a combined-cycle CT would frustrate the purpose of the source. 
Commenter 0193 noted that CAA section 111 requires EPA to develop a standard of performance that reflects the degree of emission limitation achievable through the application of BSER, taking into account the cost of achieving such reduction and any nonair quality health and environmental impact and energy requirements that the Administrator determines has been adequately demonstrated. The commenter argued that the proposed standards are not achievable for simple cycle units, are not cost effective, and fail to appropriately consider the other key factors in a BSER determination. The D.C. Circuit has found that "EPA has a statutory duty to promulgate achievable standards." See Nat'l Lime, 627 F.2d at 443. EPA is required to explain how the standard is "achievable under the range of relevant conditions which may affect the emissions to be regulated, ..." including "under most adverse conditions which can reasonably be expected to recur." See Id. at 431 n.46, 432-33. 
According to commenter 0193, the GE LMS100, the most efficient simple cycle gas turbine on the market today, would not be able to meet the proposed standard throughout its life. Indeed, it is only capable of meeting the proposed standard in a newly installed and nondegraded condition and only when operating under optimal operating conditions. The commenter provided a figure (from their comments on EPA's January 2014 NSPS), which illustrated some of the factors that affect the CO2 emissions performance of the LMS100, including ambient temperature, partial load operation, and degradation. The commenter argued that these factors show that even this highly efficient unit would be anticipated to have emissions that exceed the proposed rule's proposed limit during the lifetime of its operation. Even "as new" and under ideal condition, its operation is likely to result in an emission level of 1,020 lb CO2/MWh, which exceeds the proposed limits. 
While the LMS100's inability to meet the standard is sufficient on its own to show a lack of achievability, commenter 0193 noted that the data show that the remainder of available simple cycle units also cannot meet the standard. Indeed, the majority of simple cycle gas turbines operate at efficiencies less than 43.6 percent--the performance level necessary to reach even the upper end of the proposed range. Moreover, because simple cycle units, like combined cycle units, may be parked for some periods of time at a low load due to variability in demand, these low-load periods must also be taken into consideration in setting the standard. For example, the 43.6 percent efficiency cited for full load operation is reduced to approximately 36 percent at 50 percent load, with a corresponding increase in CO2 emissions of approximately 17 percent. Similarly, load cycling can also lead to inefficiencies. Finally, during natural gas curtailments, the units may be required to operate on fuel oil, leading to increased CO2 emissions.
Commenter 0215 urged EPA to exclude simple cycle combustion turbines explicitly from the rule because, as a practical matter, simple cycle turbines cannot achieve the proposed standards for new turbines. The commenter cited 79 Fed. Reg. at 1459 that simple cycle turbines "operate[] infrequently" and "only contribute[] small amounts to total GHG emissions," , and are vital to the reliable supply of electricity in this nation because they are a cost-effective means of providing peak power, especially during hot days. The commenter stated that simple cycle combustion turbines are also increasingly important for reliability purposes to support intermittent renewable energy sources, such as wind generation, due to their ability to quickly commence operations.  
Commenter 0257 stated that, if modified, high capacity factor, simple cycle turbines are to be subject to a GHG NSPS, the Agency should create a separate subcategory for these units, determine an appropriate BSER that does not require conversion to NGCC, analyze emissions from units in the subcategory, consider all of the requirements of section 111, and propose a limit that is achievable.

Multiple commenters (0149, 0226, 0167/0277, 0203, 0257) stated that EPA has created standards that effectively prohibit reconstruction and modification of any type of CT. The only way that an existing CT of any type can comply with the proposed standards, if it were to exceed the proposed applicability threshold, is to become a new NGCC unit. If a unit can only comply with standards for modified and reconstructed units by becoming a new unit of a different type, the proposed standards are not achievable by that unit. 

Commenters 0203 and 0260 did not agree that it is feasible to convert existing natural gas CTs to NGCC technology. Commenter 0203 believes it would be highly unlikely that NGCC is BSER for a simple-cycle turbine (and that this configuration would be technologically and economically feasible, as required by the general reconstruction rule at 40 CFR 60.15(b)(2)). This proposed requirement would also violate the statutory prohibition in section 111 against requiring any new or modified source to install and operate any particular technological system of continuous emission reduction to comply with any NSPS. EPA has failed to provide any support for concluding that conversion of existing simple-cycle units to NGCC (i.e., to achieve the BSER level of performance--modern, high efficiency NGCC technology) is technically or economically feasible across the capacity range of potentially affected simple cycle units. Somewhat casually, EPA notes that if a simple cycle CT owner/operator wishes to avoid triggering the performance standard for a project, it can simply maintain the current operational exclusion. An operational cap unaccompanied by a BSER analysis is not a lawful performance standard under the statute. 

Commenter 0260 indicated it is not technically feasible to convert existing CTs to NGCCs because there are numerous constraints at existing units, such as physical space (including the amount of land comprising the site), water availability limitations, the amount of natural gas available to the site during times of high demand, and electric power transmission restrictions. Such constraints can be avoided when building a new unit on a greenfield site, but not when modifying or reconstructing an existing unit. EPA has failed to perform any analysis of the technical feasibility of NGCC conversions at existing units. See id. at 34,989. Commenter 0260 did agree with EPA that NGCC technology with CCS is not BSER for modified or reconstructed combustion turbines. Id. at 34,989-90. An owner/operator of an existing source that is undertaking a reconstruction or modification has challenges not faced when building a new NGCC unit because the existing unit may be located at a site with space constraints that would make installation of CCS problematic. Id. at 34,989. The commenter further agreed with EPA that high efficiency simple cycle aeroderivative turbines are not BSER for modified or reconstructed combustion turbines. Id. at 34,989-90

Commenter 0255 did not support EPA's proposal that the BSER for a modified simple cycle combustion turbine is to convert the unit to NGCC unit. The commenter stated that requiring conversion of a simple cycle turbine to NGCC could create a disincentive to completing a project that might improve the efficiency of an existing simple cycle turbine. Secondly, the commenter stated that while it is technically feasible to convert some simple cycle turbines to NGCC units, it is not necessarily feasible to convert all simple cycle units to NGCC. The commenter cited constraints can preempt such a conversion, including water availability, natural gas supply constraints, existing power purchase agreements, transmission system interconnection agreements and air quality/permitting restrictions. The commenter stated that EPA should explicitly exempt simple cycle turbines from being an affected unit as long as the modification does not convert it to NGCC.
The commenter is mistaken in stating that the standard for simple cycle units is based on the performance of NGCC units. In the final rule, NGCC has not been identified as the BSER for non-base load units. NGCC has been identified as the BSER for base load units. As explained in Section IX.C.4.b of the preamble, the BSER for non-base load units and multi-fuel-fired units is the use of clean fuels. The non-base load subcategory is sufficiently flexible to allow for the continued use of simple cycle combustion turbines. See Sections IX.C.4.a.(2) and IX.C.4.b of the final rule preamble of the for more details..
Commenter 0250 stated that, because the proposed performance standards have only been demonstrated by NGCCs operating at sustained high loads, the proposal effectively dictates turbine technology choice. The manufacturer of their simple-cycle LMS100 turbines commented in support of the simple-cycle exclusion in the original (withdrawn) proposed rule, relaying that the LMS100---its most efficient simple-cycle gas turbine--is only capable of meeting combined-cycle performance standards in "newly installed and non-degraded condition and under optimal operating conditions." See Letter from Larry A. Boggs, General Electric's Director and Senior Counsel of Government Affairs, to EPA, Docket ID. EPA-HQ-OAR-2011-0660-9978 (June 25, 2012).
The EPA disagrees that the final standard has only been demonstrated for NGCC units operating at sustained high loads. An analysis of actual operating data demonstrates that the final emission standard is achievable by NGCC units in all base load operating conditions. High-efficiency simple cycle turbines like the LMS100 will be subject to the non-base load emission standard so long as they remain below the percentage electric sales threshold.
Commenter 0232 agreed with EPA that it is appropriate to base performance standards for modified and reconstructed natural gas combustion turbines on the performance of NGCC technology. The commenter noted that EPA proposed to determine that NGCC technology is the BSER for natural gas combustion turbines, for the same reasons EPA presented in the preamble to the proposed carbon pollution standards for new EGUs. As EPA observed in the preamble, NGCC is an efficient generating technology that is highly cost-effective and in widespread use. However, the commenter does not believe that a performance standard based on NGCC alone reflects the BSER for modified and reconstructed natural gas combustion turbines, as section 111 requires. Rather, the BSER for these sources consists of NGCC in addition to the requirements of an applicable section 111(d) state plan. As discussed in section I of their comments, EPA's proposed goals under section 111(d) reflect four "building blocks" that are based on well-established means for reducing carbon pollution from the power sector as a whole. The commenter strongly supported EPA's determination in the proposed emission guidelines that this system-based approach constitutes the BSER for all EGUs that were existing sources as of January 8, 2014, including existing natural gas combustion turbines.2 This system-based approach is equally effective and efficient in reducing emissions from existing EGUs that subsequently undertake modifications and reconstructions--and better fulfills the statutory criteria for BSER than any of the proposed alternatives. Indeed, failing to apply this system-based BSER to natural gas combustion turbines that modify or reconstruct would potentially lead to increased emissions if those EGUs do not also remain subject to section 111(d) state plans--a perverse outcome that would be inconsistent with the structure and purpose of section 111, as discussed in section I of their comments. Accordingly, EPA should explicitly provide in the final rule that compliance with an applicable section 111(d) state plan, together with an NGCC-based emission limitation, represents the BSER for modified and reconstructed natural gas combustion turbines. 
The commenter also urged EPA to ensure that the final standards of performance for these sources reflect the best emissions performance demonstrated by NGCC facilities. In comments filed jointly with other environmental organizations on the proposed carbon pollution standards for new EGUs, the commenter argued that they demonstrated that the proposed performance standards of 1,000 lb/MWh (for units with heat input greater than 850 MMBTU/h) and 1,100 lb/MWh (for units with heat input less than 850 MMBtu/h) could be easily achieved by almost all NGCC facilities currently in operation. More stringent standards could be cost-effectively achieved by currently available NGCC technologies, and would have substantially lower emissions.
The commenter did not take a position here as to whether EPA should adopt one of the alternative characterizations of the BSER that are presented in the proposed emission guidelines, but the commenter intended to file comments on this issue in Docket ID No. EPA-HQ-OAR-2013-0602.
 The EPA is not finalizing the proposed once-in-always-in approach. Therefore, a reconstructed base load natural gas-fired combustion turbine will have to comply with an emission standard based solely on modern, efficient NGCC technology and will no longer be an existing source subject to a 111(d) state plan.
Commenter 0192 did not disagree with EPA's conclusion that NGCC technology is BSER for natural gas turbines, but they believe the standards of performance proposed by EPA do not accurately reflect the emission limits that can be achieved by NGCC facilities under real-world operating conditions. To better reflect the emission limits that can be achieved in practice by NGCC turbines, the commenter urged EPA to revise the proposed standards for modified and reconstructed natural gas turbines.
Commenter 0214 noted that EPA proposed to require modified and reconstructed NGCC units to meet the same standards of performance proposed for new NGCCs, but argued that the improvements needed to be made at an existing NGCC unit to achieve the performance of a new NGCC unit would likely not be insignificant or cheap, pointing to EPA's statement that "the performance of combined cycle technology has improved significantly since 2000." EPA never identifies the specific technologies that an existing NGCC can implement to achieve the performance of "modern high efficiency NGCC technology." EPA also does not provide the cost of implementing these technologies at existing sources
The EPA has concluded that a reconstructed NGCC unit can achieve the same emission rate as a new NGCC unit because reconstructed units can essentially be rebuilt as new units and because the emission standard both must achieve is conservative and based on the performance of a wide variety of existing units. Therefore, the commenters are simply incorrect that the EPA has not taken real-world operating conditions into account. The specific technologies that a reconstructed turbine can use to improve efficiency are detailed in Section IX.D.3.b of the preamble.
Technical Feasibility
Multiple commenters (0146, 0158, 0189, 0193, 0203, 0212, 0239, 0250, 0257, 0267) stated that EPA's identification of NGCC technology as BSER for both modified and reconstructed natural gas-fired CTs incorrectly assumes that older units can be modified/reconstructed to achieve the same standards as new units. Referring to comments from UARG, commenter 0239 argued that the expectation that an older combustion turbine can be reconstructed to perform at the same level of a state-of-the-art new NGCC unit is flawed and therefore the proposed standards for modified and reconstructed natural gas-fired units are unachievable.  
Commenter 0187 disagreed with EPA that in all instances older natural gas combustion turbines will be able to replace sufficient hardware to become effectively a brand new NGCC unit and meet the same emission standard. The exhaust gas temperatures and flow rates of different natural gas combustion turbine makes and models can vary significantly. Not all combustion turbines are suitable for economical NGCC conversion. Furthermore, many combustion turbines may not have the space available for the equipment necessary for the NGCC conversion. NGCC technology has improved for decades and continues to improve. 
Commenter 0212 noted GTA's comments that it would not be appropriate to assume that older existing gas turbines can achieve the same levels of performance as a new state-of-the-art turbine. Existing units experience a certain amount of unavoidable degradation in performance over the gas turbine operating life. While a portion of the degradation can be recovered as a part of the turbine upgrade, it is never feasible to recover all of the performance to bring the unit to the same level of performance as a new, clean gas turbine installation. 
Commenter 0203 questioned whether an existing source, with inherently fewer degrees of engineering (design and construction) freedom, can meet the same emissions standard as a new, greenfield source. In the absence of specific emissions data or case studies, any presumption regarding the appropriate standard should be in favor of setting a higher (i.e., less stringent) standard for reconstructed units than those which are new. The commenter also noted that all equipment degrades over time and becomes less efficient. If EPA sets the performance standards too stringently, sources that trigger a NSPS (modifications or reconstructions) may have no choice but to shut down rather than move forward with a triggering project. The commenter did not believe it was the intent of Congress when it passed section 111 (or EPA when it promulgated 40 CFR part 60) to force shutdowns of power plants due to NSPS. EPA should be careful to avoid this unintended consequence of NSPS set too stringently. Standards must be achievable under section 111.
Commenter 0267 stated that the proposed standards for new units are based on ideal operating conditions that do not reflect the range of real world operating conditions nor the likely in-use degradation that is experienced by every single plant when operating for long periods of time. Accordingly, the standards are unlikely to be achievable in the real world. Applying these same standards to modified and reconstructed units only compounds the problem. Existing units that trigger modification or reconstruction will be older units that are based on earlier designs that do not incorporate the design improvements that have made today's turbines highly efficient. While modification and reconstruction projects often improve performance, they generally cannot restore the unit to the performance on par with a new unit. Existing units that undergo modification or reconstruction will also generally not be able to achieve the same standards as newly constructed units due to unrecoverable losses in efficiencies from degradation.
Commenter 0257 stated that setting the same emission standards for both new and modified and reconstructed sources is anomalous. As a technical matter, older units are not capable of replacing sufficient equipment to match the emissions performance of a brand-new NGCC unit on a cost-effective basis, and the Agency has not provided any technical support for basing its proposed standards on such an assumption. New sources may have several inherent advantages over existing, yet-to-be-modified or reconstructed sources that would make less aggressive standards for modified and reconstructed sources reasonable. New sources are constructed so that all component pa1ts are integrated from the start. Sources that undergo modifications or reconstructions, by contrast, must contend with the added expense and technical hurdles of adapting new technology to existing infrastructure. Existing, yet-to-be-modified or reconstructed sources, by contrast, must work within the confines of their existing sites.
Commenter 0257 noted that an industry group consultant has analyzed emissions data from 248 NGCC turbines that began operation between 2007 and 2013. The consultant found that about 90 percent could meet a limit of 1,000 lb CO2/MWh, while 98 percent of large (> 850 MBtu/hr) units could meet a CO2 limit of 1,100 lb CO2/MWh. These results are substantially below the 99 percent achievability target that EPA has historically applied in NSPS rulemakings and would require a significant number of NGCC units to retire if they became subject to the proposed standards. In order for 99 percent of turbines to comply, EPA would need to finalize an NSPS of over 1,400 lb CO2/MWh. The proposed NSPS for modified and reconstructed units would apply to any existing combustion turbine, including many that commenced operations before 2000. EPA has provided no information to justify that modified units constructed in the 1990s can match the emissions performance of new turbines with more efficient designs that were built up to 25 years later. These older units have no available means to improve heat rate beyond design levels. Older turbines cannot replicate the emissions perfom1ance of modern turbines through overhauls or other refurbishments. 
Commenter 0189 argued that applying these stringent limits to existing units that are modified and reconstructed is not appropriate since it would be more difficult to incorporate retrofits to existing units. There is no add-on pollution control technology that is feasible at this time to reduce CO2 emissions from modified and reconstructed units, such as would be available for compliance with a NOx performance standard. The EPA has not provided justification that these proposed CO2 limits are achievable for modified and reconstructed units. For example, the commenter has considered as a resource generation option the conversion of an existing simple cycle natural gas-fired combustion turbine to a combined cycle at some future time. This conversion would serve to lower the current maximum load emission rate of the unit which is approximately 1,300 lb CO2/MWh to between 1,000 and 1,100 lb CO2/MWh, making it more efficient. However , the modified unit could be subject to the more stringent limit in the proposed rule of 1,000 lb CO2/MWh for sources with heat input >850 MMBtu/h. Since this modified unit would not be able to comply with the EPA's proposed limit, they may not be able to utilize this simple cycle conversion to combined cycle as a resource option to meet growing demand. 
Commenter 0146 disagreed that an NGCC unit that is 20 or more years old will invariably be able to undertake sufficient equipment upgrades to meet the same CO2 emission rate as modem technology. To the contrary, as noted in comments of UARG, due to changes over time in the design efficiency of NGCC units, it is not possible to bring older combustion turbines up to the efficiency levels of modem units. By design, newer NGCC units are capable of operating at significantly higher efficiency than older units in the existing NGCC fleet - particularly those built before the year 2000. "NGCC technology," as EPA terms it, has been evolving and improving for decades and continues to improve. Most newer turbines utilize new high-temperature materials that allow them to achieve higher efficiencies that are not attainable for older turbines using other, lower-temperature technologies. See Lowell L. Smith, "Analysis of the Proposed Modified and Reconstructed GHG NSPS for Natural Gas Combined Cycle Turbines," (Oct. 7, 2014).
Although EPA is always required to ensure that any NSPS is achievable for the entire source category, commenter 0146 stated that it is particularly important that EPA provide a sufficient margin for achievability here, given that the Agency itself acknowledges that these units have no other feasible means to lower their CO2 emissions profiles beyond the rate associated with each unit's "NGCC technology." Therefore, an older NGCC unit that is modified or reconstructed but cannot meet the proposed NSPS with its current NGCC configuration does not have an option of installing add-on technology and would either have to entirely replace its existing facility with a new modem turbine (and likely a new heat recovery steam generator) at exorbitant cost or retire altogether in order to comply, as noted in the Smith report. An NSPS for these units that is not achievable could place a significant portion of the existing generation capacity at risk of forced retirement. The commenter owns two NGCC facilities that were constructed in 1986 and 1994, and if the units at either of these facilities were deemed to be subject to the proposal, those units would be unable to meet a standard based on the emissions performance of NGCC units constructed 20 or more years later. Commenter 0158 argued that section 111(b) standards for CTs must be achievable at the time of proposal.
Commenter 0193 noted that they are continually advancing gas turbine technology, investing nearly a half billion dollars per year in gas turbine development, with the primary objective of improving turbine and plant level efficiency. As a result of these significant investments, today's new gas turbines are more efficient than the existing fleet of gas turbines. Through this ongoing technology investment, while a 7B gas turbine manufactured in the 1970s generates roughly 60 MWs with an efficiency of roughly 30.5%, today's state of the art 7HA gas turbine can generate 330 MWs with an efficiency in excess of 41%. While many of the latest technology advancements can be retrofitted to improve the performance of existing turbines, the commenter argued it is simply not feasible to bring the existing fleet of turbines to the same advanced level of performance as a new NGCC plant installed today. Even when retrofitted, existing plants will not share the operating capabilities of today's NGCC plants. This includes the fast start capabilities built into today's plants. Today's plants are able to start faster than a plant built even a decade ago. This faster start capability means less CO2 emissions during start events leading to a lower 12-month average emission rate. Heat recovery boilers, and other systems, must be specially designed to accommodate the increased thermal stresses resulting from these faster starts. These are capabilities that cannot be retrofitted into an existing boiler. Instead, these features must be designed into the equipment during initial installation. Because of these factors and the unavoidable performance degradation encountered by all operating plants, existing plants will have a higher 12-month rolling average CO2/MWh emission rate than a new plant installed today.
Commenter 0193 noted that, because modified and reconstructed units are by definition existing units, their base level of performance is generally less efficient than a comparable new unit due to non-recoverable sources of in-use degradation and the continued improvements in the design of new units. While the modification or reconstruction may help bridge this efficiency gap, in most instances, it is unlikely to completely change the existing unit so that it operates like a newly constructed turbine. Moreover, EPA based its January 2014 BSER determination for new units on emissions data from 2007 to 2011 for natural gas-fired (non-CHP) combined cycle units that commenced operation on or after January 1, 2000, and that reported complete electric generation data, including output from the steam turbine, to EPA. This database, however, does not include all existing gas turbines, many of which went into service prior to 2000.
Commenter 0250 noted that the proposed NSPS for GHG emissions from gas-fired stationary CTs established two subcategories: those with heat input greater than 850 MMBtu/h, and those with less. However, no data in the administrative record demonstrates that stationary CTs can meet the proposed standard for either category. This is because the proposed standards are "based on modern, efficient natural gas combined cycle technology." The Technical Support Document for the New NSPS explains that EPA "reviewed the emissions rates of natural gas-fired combined cycle units" to determine that "over 96% of these units achieve the proposed standards with no excess emissions." See Memorandum from EPA to the EGU NSPS Docket, Docket ID No. EPA-HQ-OAR-2013-0495-0082 (Sept. 2013). EPA only found that "because over 90 percent of small and large existing [combined-cycle] facilities are currently operating below the emission rates of 1,100 lb CO2/MWh and 1,000 lb CO2/MWh, respectively, these rates are considered BSER for new [combined-cycle] facilities in those respective subcategories [and] [t]hese values represent the emission rates that a modern high efficiency [combined-cycle] facility located in the U.S. would be able to maintain over its life." EPA made no such finding for stationary CTs. Accordingly, EPA must set other performance standards, considerate of the differences in thermodynamic performance, operating characteristics, and emissions profiles not only between NGCCs and SCTs, but also among SCTs. The standards could even consider that aeroderivative SCTs are more efficient at low loads.
Commenter 0215 stated that it is contrary to general engineering principles that an existing source, with inherently fewer degrees of engineering freedom, could meet the same emission standard as a new, greenfield source. The commenter stated that reconstructed NGCC units should have a standard that is less stringent than new units.  
In order to demonstrate that its proposed NSPS for reconstructed and modified units are achievable, multiple commenters (0146, 0149, 0150, 0189, 0193, 0199, 0201, 0203, 0257, 0267) stated that EPA must take into account a number of important factors. The commenters stated that EPA must "(1) identify variable conditions that might contribute to the amount of expected emissions, and (2) establish that the test data relied on by the agency are representative of potential industry-wide performance, given the range of variables that affect the achievability of the standard." See Sierra Club, 657 F.2d at 377 (citing Nat'l Lime Ass'n, 627 F.2d at 433). In other words, the validity of EPA's achievability determination depends on how fully it has accounted for the variations among sources in the regulated category that could affect emission levels. 
Commenter 0146 noted that courts have repeatedly rejected NSPS that EPA deemed "achievable" based on test data from a narrow set of sample sources that did not represent the full range of relevant variability among sources to which the standard will apply. See e.g., Nat'l Lime Ass'n, 627 F.2d at 432-33; Portland Cement Ass'n, 486 F.2d at 396, 402. In addition, the commenter noted that EPA has traditionally set NSPS such that at least 95 percent of units subject to the standard would be expected to comply. The commenter stated that this reflects the fact that an NSPS "establishes what every source can achieve, not the best that a source could do." See Letter from Gary Mccutchen, Chief, New Source Review Section, EPA OAQPS, to Richard E. Grusnick, Chief, Air Division, Ala. Dep't of Envtl. Mgmt. at 1 (July 28, 1987) ("Mccutchen Letter"). Commenters 0150 and 0201 went even further. Commenter 0150 argued that accounting for variable conditions generally requires a showing that at least 99 percent of new units would be expected to be able to comply with the NSPS. Commenter 0201 argued that EPA should modify the proposed standard to account for the operational characteristics of the entire gas generation fleet and address the inherent engineering limitations in modifying or reconstructing an existing unit as compared to building a new unit. Commenter 0201 said that this would address the issue where stationary CTs may not be able to meet the current proposed standards under certain operating conditions. The commenter noted that there are instances where stationary combustion turbines are needed to support large amounts of intermittent renewable energy generation, which requires the units to cycle on and off with much more frequency than a plant used to support base load energy needs. A stationary combustion turbine that has frequent startup and shutdown periods will have higher CO2 emissions than a unit that operates at steady state over the same time frame. In a similar comment, commenter 0149 stated that, in order to be achievable, the standards must reflect all operating conditions and anticipate that, in the future, NGCC units will be needed to help integrate increasing amounts of variable renewable generation into the grid. EPA's proposed section 111(d) guidelines, which contemplate a dramatic increase in the deployment of wind and solar generation, underscore these concerns. As more renewables are brought onto the system, NGCC units will ramp up and down more frequently, or, will run at lower capacity factors generally, which will degrade heat rates and increase emission rates.
Commenters 0193 and 0267 stated that EPA must assure that the final standards for new and modified or reconstructed gas turbines are achievable over the lifetime of the facilities considering the range of conditions, dispatching needs, capacity loads, in-use degradation, and startup, shutdown, and malfunction. The commenters noted that the Sierra Club courts has concluded that the standards must be based on reliable information representative of "potential industry-wide performance, given the range of variables that affect achievability of the standard." See Sierra Club v. Costle, 657 F.2d 298, 377 (D.C. Cir. 1981) (citing Nat'l Lime Ass'n v. EPA, 627 F.2d 416, 433 (D.C. Cir. 1980). EPA cannot base the standard on assumed operating conditions that are not reflective of how plants operate in practice, and EPA has not justified in the record a basis for dictating changes to the way that operators ensure their units are ready to come on-line in the event they are needed or ramp down when demand is reduced. EPA should revise the standard to ensure that it is achievable for the "industry as a whole" and "for all variations of operating conditions being considered anywhere in the country." See Nat'l Lime Ass'n. The performance of turbines in use is affected by a number of factors, including a range of ambient temperature, water cooling needs and technologies, operating demands, including increased load variations, and in-use degradation. Including startup and shutdown emissions in the standards will also further challenge units in complying with any new standard.
Commenter 0193 noted that, when newly installed, their fleet of gas turbine products is capable of achieving the proposed emission limitations at base load operations. Establishing standards based on these optimal conditions, however, does not reflect the real world operating conditions or account for everyday scenarios where an NGCC unit will not be able to achieve the proposed limits. As a result, EPA's proposed limits do not represent BSER for these units. According to the commenter, section 111(a)(1) of the CAA requires EPA to establish standards of performance that reflect "the degree of emission limitation achievable through the application of the best system of emission reduction which (taking into account the cost of achieving such reduction and any nonair quality health and environmental impact and energy requirements) the Administrator determines has been adequately demonstrated." See 42 U.S.C. 7411(a)(1). In determining what is achievable and adequately demonstrated, EPA must consider the range of current and anticipated operating conditions in the United States. This includes the challenges faced by dramatically different physical environments and operational needs in response to rapidly changing electricity demands. Any standard that fails to account for these real world conditions does not represent BSER because it is not achievable or adequately demonstrated. EPA must take into account the following important factors in establishing BSER for new units. 
 Steam turbine condenser cooling: In a NGCC unit, condensing the steam back down to water creates a vacuum at the exit of the steam turbine that greatly enhances the steam cycle efficiency. The condensing of the steam, however, requires a cooling source, which was achieved historically by once-through cooling water drawn and returned to a body of water. Given that once-through cooling is now prohibited on new plants in the United States, most plants today rely on a cooling tower, or air cooling using an air-cooled condenser. Cooling towers lose a portion of the cooling water through evaporation. At sites, notably in arid regions, where water availability is limited, air-cooled condensers are used. Unfortunately, air-cooled condensers are not as efficient at creating the vacuum at the exit of the steam turbine resulting in a decrease of plant efficiency. As a result, plants with air-cooled condensers can be up to 1 percent less efficient than those with cooling towers at nominal conditions.
 Ambient temperatures: A gas turbine's energy production and efficiency depend on the mass of air that can pass through the turbine. On hot days, air density is lower so less air mass can pass through the machine, making the turbine less efficient on hot days compared to cold ones. A nominal gas turbine will be 0.5 percent less efficient on a 100-degree day than a 59- degree International Organization for Standardization (ISO) day. Hot temperatures also affect the performance of the air-cooled condensers noted above. A plant with a cooling tower will be approximately 4 percent less efficient on a 100-degree day than a 59-degree ISO day, while a plant with an air cooled condenser will be closer to 8 percent less efficient. Thus, a plant located outside of Las Vegas, with an air-cooled condenser to limit water usage, could easily have 5 percent higher CO2 emissions than the fleet average on hot days. As a result, a plant in a hot arid climate will be significantly less efficient than a plant in a cooler climate with available water resources. Because of these factors, the emission level must be established with sufficient margin to accommodate these temperature and design driven efficiency differences.
 Varying Operational Loads: Under the proposed standard, compliance is determined on a 12-month rolling sum, including all potential operation during that period. NGCC plants are most efficient when operating at base load during normal operating conditions. While EPA might assume that operators can dictate and adjust their operations to achieve the standard or to operate in a particular mode, like base load, this is not the case. A large proportion of the installed fleet is operated under an approach referred to as Automatic Generation Control (AGC). In AGC mode, the grid operators, not the plant, effectively control plant operation and load by sending demand signals to the plant with the plant automatically adjusting to satisfy that demand requirement. Operational flexibility is a core gas turbine attribute. Many plants do not operate consistently at base load, but rather may be "parked" at low load during overnight periods. The reason for parking these units is to reduce production during times of low demand (e.g., when people are sleeping and the lights are out). A parked unit will have higher emissions per MWh but overall lower mass emissions (and fuel savings) compared with maintaining base load. A BSER determination requires EPA to consider these factors in setting the standard for new units. Moreover, any BSER analysis should also take into account expected changes in operations to support increased renewable energy. It is expected that grid operators soon will require increased operating periods at a part-load or even minimum-load "parked" condition. This condition allows the unit to be on "hot standby," thereby enabling the plant to respond rapidly to load changes without the time delay and equipment stresses of full plant shutdown and startup sequence. The rule must accommodate these future trends in operation.
 Start-up and Shutdown Emissions: During a plant startup or shutdown, there are periods of fuel burn during which either no energy is being produced or relatively low levels of energy are being dispatched to the grid. As a result, during startup and shutdown, the mass of CO2 emitted per MWh produced is very high and will certainly exceed 1,000 lb. CO2/MWh limit. Emissions during startup are also variable between installations and can vary even at the same plant due to their dependence on site- and time-specific factors, such as equipment temperature at the initiation of startup. For example, while a plant that has been shut down overnight can start in generally less than an hour because the equipment is still warm from previous operation, that same plant may require over three hours to achieve full operating temperature if it has been down for a week and the equipment is at ambient temperatures. A rapid startup can induce significant thermal stresses to the equipment, especially "cold" equipment, which means that the startup period must necessarily be longer. This will result in significantly higher CO2 emissions. The more startups and shutdowns and the more cold starts, the greater the challenge a given unit will have in complying even under otherwise optimal conditions.
 Supplemental Duct Firing: Supplemental duct firing is commonly deployed in NGCC plants to increase steam production and boost the plant output during periods of peak demand. Supplemental thermal energy is introduced in the ductwork leading into the heat recovery steam generator (HRSG) through the use of gas combustors, commonly referred to as duct burners. This technology provides essential inexpensive capacity, especially on hot days with peak energy demand. The use of supplemental firing to provide peak generation, however, reduces overall plant efficiency with a resulting increase in CO2/MWh.
 Combined Cycle Power Plant Degradation: Various operational and environmental factors will cause NGCC plant performance to degrade over time. Representative causes of performance degradation are fouling of the gas turbine compressor, wear of flow seals allowing leakage within the gas turbine and steam turbine flow paths, erosion and corrosion of heat exchanger surfaces in the HRSG, steam condenser and cooling tower, and other performance impacts to ancillary components. This is not an exhaustive list but rather illustrates typical causes of unavoidable performance degradation that will be encountered at any NGCC plant. While some degradation is partially recoverable through maintenance, some degradation is permanent. The most significant cause of power plant degradation is fouling of the gas turbine compressor. The rates of both compressor fouling and performance loss are a result of the variation in environmental conditions, machine operating scenarios, and maintenance practices. Performance loss during normal operation is minimized by periodic on-line compressor water washes, in which water and cleaning agents are sprayed into the compressor while the turbine is in operation, and off-line compressor water washes, in which water and cleaning agents are sprayed into the compressor while the unit is not operational, allowing a more thorough compressor cleaning. Annual, average plant efficiency would form the best basis to compare degradation effects to the 12-month average CO2 emission standard. The level of degradation experienced by any individual plant varies greatly depending on site conditions, operating level, start frequency, and ambient factors--such as to sea water and associated sea salts in the air. Given the wide range of potential degradation causes, it is difficult to establish a representative value, although recent permit BACT evaluations have included degradation margins as high as 5 to 10 percent to account for the broad range of operational effects on NGCC plants.
Thus, in establishing BSER for new, modified, and reconstructed units, commenter 0193 argued that EPA must consider the impact of all of these factors on the likely in-use performance of new gas turbines over time. While the commenter's fleet of gas turbines can readily achieve the proposed emission standards when operating new at base load steady-state conditions, they cannot meet the standards if they are used as expected to meet the range of power demands needed in today's electricity market. EPA cannot ignore these real world factors and operating conditions. Failing to account for these conditions would be inconsistent with CAA section 111, and would undermine the rapid deployment of cleaner energy sources, including renewable sources. EPA should revisit its approach and develop a final action that is supported by the record.
In a similar comment, commenter 0257 questioned whether the proposed emissions standard for NGCCs has been adequately demonstrated to be achievable continuously under normal, real-world operating conditions. The inability to meet the 1,000 lb/MWh proposed standard can be attributed to several factors. For example, the efficiency of an NGCC unit and the CO2 emissions rate are affected by the following factors:
 Operations at part load: Emissions (on an output basis) increase significantly as units are operated at partial load for a variety of reasons. Many NGCC units, even those designated as "baseload," are required to operate at lower capacities to accommodate integration of renewables and other system, grid or electricity market specific needs. Additionally, combined-cycle units that have a two-on-one or three-on-one configuration may have occasions to operate with one CT out of service for maintenance, which reduces efficiency due to reduced heat input into the steam cycle. And as the renewables that are mandated by this rule are integrated into the grid, lower and interrupted load service can be expected for NGCCs, especially the older, smaller units that are not as efficient, making their emission rates even higher.
 Use of backup fuel: The proposed standard does not appear to incorporate the use of oil as a backup fuel. The use of oil in times of natural gas curtailment (e.g., severe weather events, such as hurricanes or interrupted natural gas supply) will increase the average CO2 emissions rate. This was vividly demonstrated during the very cold winter of 2013-2014 when natural gas was in short supply due to pipeline constraints and oil was called on to allow CTs to supply the maximum load to the electric grid. Furthermore, given the number of NGCCs predicted as the primary technology of choice for new power generation and their potential impact on natural gas supply and natural gas infrastructure, companies may be required by state commissions or siting authorities to have oil available as a back-up fuel in order to support electric reliability.
 Startup and shutdowns: These actions are part of normal operation but do not constitute baseload operation and should be exempt from the rule. At start-up, a combustion turbine operates essentially as a simple-cycle unit until the steam generator is heated. This is part of the design of an NGCC unit. Some NGCC units do not simply follow load, they are taken off -line and restarted on a daily basis, increasing the CO2 emission rate based on market conditions.
 Performance degradation: While EPA states that it accounted for performance (efficiency/heat rate) degradation in setting the proposed standard (based on the design performance specification analysis), the standards do not account for the fact that each unit is specific in design (e.g., number and efficiency of inlet filters) and operating site conditions (dusty open plants, high ambient temperature), and the fact that the percent degradation over time is variable. Referring to the "What You Should Know About Design Ratings" document included in the 2012 proposal docket10, ''power and heat rate may degrade by 2-6 percentage points during the first 24,000 hours of operation." To the extent that some degradation can be recovered, it is not always possible or feasible to correct a slightly degraded function immediately, as repairs often involve extended tear downs and must be scheduled well in advance. 
 Multiple commenters (0142, 0146, 0150, 0152, 0203, 0211, 0215, 0226, 0257, 0260) argued that EPA had failed to provide the data and analysis demonstrating that the proposed standards are achievable by modified and reconstructed units, or explaining how the Agency reached its conclusions. Commenter 0211 stated that EPA had not supported its conclusion that modified and reconstructed simple cycle units can meet the same standard as NGCC units. Commenter 0211 stated that EPA posited, without supporting technical information, that simple cycle turbines can be modified and reconstructed to become NGCC units. Commenters 0211 and 0142 stated that a search of the docket does not reveal any TSD for modified or reconstructed natural gas-fired combustion turbines, despite such being referenced in the preamble to the proposed rule (Commenter 0211 stated that EPA does include a TSD in the docket regarding BSER for Reconstructed Steam Generating Units and IGCC facilities and a TSD for the New Source GHG NSPS Combustion Turbine Standard, however, the document does not directly refer to or consider modified or reconstructed sources). Commenter 0211 stated that no economic analysis is provided to support the assertion that when operated at higher capacity factors, conversion of a simple cycle unit to an NGCC is cost-effective. While the commenter 0211 believes that cost savings may be possible given the relative efficiency of NGCC units, it was not clear that this would be true in every instance and, in any event, EPA offered only unsupported, conclusory statements on this matter. Commenter 0142 stated that without such missing data and related materials, States and the public cannot properly determine the basis on which the EPA claims that these emission standards are achievable or reasonable. 
In a similar comment, commenter 0203 also stated that the docket does not contain any TSD supporting the proposed NSPS for reconstructed subpart KKKK units, but contains only a TSD corresponding to the January 2014 proposed NSPS for new units. Commenter 0203 stated that no explanation as to why this TSD for new units justifies or is relevant to the proposed performance standards for reconstructed or modified units was given. Commenter 0203 stated that EPA's analysis of emissions in this TSD in support of its new NGCC unit proposal was restricted to CO2 emissions data from 2007 to 2011 for NGCC turbines; data from earlier years for older units was ignored. Because the proposal addresses the expected performance of existing units (which are then modified or reconstructed), however, commenter 0203 believes data for older units must also be included (and properly addressed) in any analysis that determines the capabilities of a reconstructed (or modified) unit. Commenter 0203 also noted that EPA failed to provide the emissions data that justify the proposed standard and that account for the many variables that affect examined emission rates, and without such data, EPA has failed to show that its proposed standards are achievable, in derogation of CAA section 307(d)(3).
Similarly, commenter 0257 stated that EPA had not provided any TSD or other information to justify that modified units that were constructed in the 1990s can match the emissions performance of new turbines with more efficient designs that were built up to 25 years later. Commenter 0257 requested that EPA set a limit whereby at least 99 percent of modified units would be expected to comply, consistent with previous NSPS determinations.
Commenter 0150 and 0215 stated that EPA had failed to provide data or analysis in the record explaining why EPA believes the emission rates for modified and reconstructed units are achievable under the range of operating conditions these units will experience. Commenter 0150 stated that  EPA is required to "(1) identify variable conditions that might contribute to the amount of expected emissions, and (2) establish that the test data relied on by the agency are representative of potential industry-wide performance, given the range of variables that affect the achievability of the standard." See Sierra Club v. Costle, 657 F.2d 298, 377 (1981) (citing Nat'l Lime Ass'n, 627 F.2d 416, 433 (D.C. Cir. 1980)). Commenter 0150 stated that This generally requires a showing that at least 99 percent of new units would be expected to be able to comply with the NSPS, and absent that information, the standards, if finalized, would necessarily be arbitrary and capricious.
Commenter 0215 stated that an analysis of existing units performed by a UARG consultant demonstrates that a substantial number would likely not be able to achieve the Proposed Standards. 
Multiple commenters (0149, 0161, 0167/0277, 0193, 0195, 0201, 0214, 0261) argued that the standards for modified and reconstructed natural gas CTs should not be the same as the standards for new sources. Commenter 0201 stated that a modified and reconstructed unit is different than a brand new one because there are inherent limitations on the modifications that can be performed on an existing unit. Commenter 0161 stated that the inherent differences between new and modified/reconstructed plants justify a higher emission standard for the latter. According to commenter 0161, when an existing plant is modified or reconstructed, the process is constrained by the layout of the facility, siting issues, compatibility problems between new and existing equipment, and other challenges. Commenter 0161 stated that those types of problems do not exist when constructing a new plant. Commenter 0161 stated that EPA recognized the important differences between new and modified/reconstructed units when it set the emission standards for modified and reconstructed coal-fired units. Commenter 0161 cited the New Source Proposal, that the emission standard for new coal-fired units is 1,100 lbs CO2/MWh over a 12-month operating period or 1,000-1,050 lbs CO2/MWh over an 84-month operating period. Commenter 0161 stated that under the modified/reconstructed source proposal, the emission standards for reconstructed coal-fired units are 1,900 lbs CO2/MWh for sources exceeding 2,000 MMBtu/h and 2,100 lbs CO2/MWh for sources with a heat input rating of 2,000 MMBtu/h or less. Commenter 0161 stated that the modified source proposal establishes less stringent emission standards for modified and reconstructed coal-fired units, as compared to new coal-fired units, and that EPA should adopt the same approach for modified and reconstructed natural gas-fired units. 
Commenters 0149, 0167 and 0277 stated that setting the same emission standards for both new and modified and reconstructed sources is anomalous on its face. Commenters 0149 and 0167/0277 stated that new sources may have several inherent advantages over existing, yet-to-be-modified or reconstructed sources that would make less aggressive standards for modified and reconstructed sources reasonable, and that EPA has recognized this. Commenters 0149 and 0167/0277 stated that new sources are constructed so that all component parts are integrated from the start, while sources that undergo modifications or reconstructions, by contrast, must contend with the added expense and technical hurdles of adapting new technology to existing infrastructure. Commenters 0149 and 0167/0277 stated that the potential achievability, costs and energy impacts of the standards may differ dramatically for modified and reconstructed sources compared to new sources. Commenters 0149 and 0167/0277 stated that new greenfield development projects can be sited on plots of land of sufficient size to accommodate required emission control technology, while existing, yet-to-be-modified or reconstructed sources, by contrast, must work within the confines of their existing sites. Commenters 0149, 0167 and 0277 stated that EPA nowhere indicates that it considered these and other distinguishing factors in proposing the same emission standards for new sources as well as modified and reconstructed sources.
Commenter 0193 stated that important differences remain between newly constructed units and existing units that are modified and/or reconstructed. Commenter 0193 stated that existing units, even when reconstructed, face greater physical and technical constraints than new units and existing units may trigger modification in response to a number of different physical or operational changes. Commenter 0193 stated that these differences should influence the BSER determination and the resulting standards for modified and reconstructed units. Commenter 0193 stated that while EPA correctly requests comment on many of the issues raised in the proposed rule for new units regarding the emission standards, neither the statute nor real-world considerations require that new and modified and reconstructed unit have the same standards.
Commenter 0195 stated that it is inappropriate for EPA to force modified and reconstructed combustion turbines to comply with the same standards as those for new units due to technical constraints. Commenter 0193 stated that for newly designed units, manufacturers have access to computer modeling that allows for the optimization of the turbine and auxiliary systems design so as to assure development of the most efficient unit, whereas modified and reconstructed units cannot take advantage of the same design synergies that would allow new units to meet such stringent limits.
Commenter 0261 stated that reconstructed combustion turbines should not be forced to comply with the same standards as those for new units given that existing units were not designed with the technical standards and technology that apply to new units today.
Commenter 0214 stated that modified and reconstructed NGCC units and simple-cycle CTs would not be able to continuously meet the same performance levels as newly constructed units that maintain state-of-the-art technologies. Commenter 0124 stated that EPA has provided no factual support for its contrary assertion and should withdraw its proposal to require modified and reconstructed NGCC units and simple-cycle CTs to reach the same levels of performance as new NGCC units.
Multiple commenters (0150, 0192, 0193, 0195, 0224, 0242, 0253, 0259, 0267) noted that the Agency proposed standards of performance for modified/reconstructed units of 1,000 lb CO2/MWh-gross for large (greater than 850 MMBtu/hr) units and 1,100 lb CO2/MWh-gross for small (less than 850 MMBtu/hr) units. The commenter stated that these standards are identical to EPA's proposed standards for new units. Commenters also noted that EPA requested comment on a range of limits for modified/reconstructed stationary combustion turbines: for large units, a range of 950-1,100 lb CO2/MWh-gross; for small units, a range of 1,000-1,200 lb CO2/MWh-gross. Commenters indicated that the standards for modified/reconstructed stationary combustion turbines should be increased to 1,100 and 1,200 lb CO2/MWh, or higher, for large and small units, respectively. 
Commenter 0150 argued that existing sources are not likely to be capable of meeting the same emission rates as new sources, and that EPA has not provided a basis for proposing otherwise. Commenter 0150 stated that unless EPA can provide technical information to the contrary, reconstructed NGCC units should have a standard at least 100 lb CO2/MWh-gross higher than new units. Commenter 0150 believes that the EPA failed to identify what additional "headroom" the NGCC would need where those units are intermittent (mostly wind- following) if the state has existing or new renewable energy requirements. Commenter 0150 stated that as NGCC units age, they lose efficiency. Commenter 0150 stated that NGCC units that engage in heavy ramping for wind following may actually increase their CO2 emissions as those NGCC units age or as more intermittent renewables are added. The commenter stated that APPA does not know what that headroom should be but that higher CO2 emissions should be anticipated as the modified/reconstructed units age.
Commenter 0193 noted that EPA based its January 2014 BSER determination for new units on emissions data from 2007 to 2011 for natural gas-fired (non-CHP) combined cycle units that commenced operation on or after January 1, 2000, and that reported complete electric generation data, including output from the steam turbine, to EPA. The commenter stated that this database does not include all existing gas turbines, many of which went into service prior to 2000, and as a result, a BSER determination requires EPA to allow up to a 10 percent higher allowable emission rate for modified and reconstructed units than the achievable BSER standard of 1,100 lb/1,200 lb CO2/MWh-gross standard for new gas turbines.
Commenters 0195 and 0224/0253 recommended EPA set limits for modified/reconstructed sources at the high end of the range for both the large and small categories, while commenter 0259 recommended EPA finalize emission limits of 1,200 lb CO2/MWh-gross for large combustion turbines and 1,300 lb CO2/MWh-gross for smaller combustion turbines (At least one of the CTs in commenter 0259's fleet that has been installed in the last ten years cannot meet the currently proposed number for reconstructed CTs). Commenters 0195 and 0259 stated that, for newly designed units, manufacturers have access to computer modeling that allows for the optimization of the turbine and auxiliary systems design so as to assure development of the most efficient unit. Commenters 0195 and 0259 stated that modified and reconstructed units cannot take advantage of the same design synergies that would allow new units to meet such stringent limits. Commenters 0195 and 0259 stated that given these technical limitations, EPA's standards for gas-fired stationary CTs should differentiate between new, modified, and reconstructed units.
Commenter 0224/0253 noted that modified and reconstructed units are constrained by the existing unit's design and configuration, while newly built units are not so constrained. Commenter 0224/0253 stated that consequently, modified and reconstructed units cannot achieve as stringent a level of emission reductions. Commenter 0224/0253 noted that EPA has consistently subjected existing sources to less stringent standards than new sources.   Commenter 0224/0253 stated that applying standards at the high end of a reasonable range of emission limits for reconstructed and modified sources would provide the same leeway, recognizing that the technical and design constraints of existing sources limit options for reconstruction or modification. Commenter 0224/0253 stated that as units get older, they do not run as efficiently as when they were new, which diminishes their ability to achieve as stringent emission levels as new units. Commenter 0224/0253 stated because EPA has established a range of reasonable emission limits and setting a limit at the high end still falls within the reasonable range, adopting the high end of the range for modified or reconstructed EGUs is permissible.
Commenter 0192 believes that EPA should increase the standard of performance for both large and small natural gas turbines by 100 lb CO2/MWh, to 1,200 lb CO2/MWh (HHV) for units less than 1500 MMBtu/h and to 1,100 lb CO2/MWh (HHV) for larger units. Commenter 0192 stated that these higher standards of performance would better reflect the emission limits that can be achieved by natural gas turbines in real-world conditions. Commenter 0192 stated that NGCC turbines operate most efficiently when providing baseload power; however, in the real world NGCC turbines more commonly provide intermediate load power where operations are less efficient and emissions rates increase. Commenter 0192 stated that to account for inefficiency created by more frequent cycling due to intermediate loads, along with steam turbine condenser cooling, changes in ambient temperature, supplemental duct firing, and plant degradation over time, it is appropriate to raise the proposed standards of performance by 100 lb CO2/MWh.
Commenter 0267 argued that, given the age distribution of the existing fleet and what turbines are likely to achieve in-use over the range of real world operating conditions, BSER for modified and reconstructed turbines is an emission limitation of 1,200 lb CO2/MWh for larger units with a heat input rate above 1,500 MMBtu/h (HHV) and 1,300 lb CO2/MWh for smaller units with a heat input rate equal to or less than 1,500 MMBtu (HHV).
Commenter 0242 did not feel meaningful comments could be formed due to the many variables and ranges presented...all compounded by the multitude of changes being considered regarding applicability. Commenter 0242 stated that while it is apparent the lower end of the range is not achievable, it may be possible the higher end of the range could be achievable for some types of units. Commenter 0242 stated that EPA had not provided enough specificity for the commenter to determine achievability of the proposed standards of performance and provide meaningful comment.
Commenter 0209 requested that EPA increase the emissions limit for modified and reconstructed NGCC units from 1,000 or 1,100 lb CO2/MWh to 1,250 lb CO2/MWh to account for load following. The commenter stated that while a new NGCC unit would likely meet an emission rate limit of 1,000 lb CO2/MWh when utilized for base load generation, it is unclear whether NGCC units can sustain that level of efficiency when they are used for load following, such as to firm up delivery of wind and other intermittent generation. The commenter stated that the Front Range Power Plant is an NGCC plant owned and operated by the commenter's utility that began operation in 2002. The commenter stated that analysis of the historic emissions data from the part 75 certified CEMS demonstrates rolling 12-month emission rates above 1,000 lb CO2/MWh for a number of months in 2011. The commenter stated that these slightly higher emissions correspond with lower generation months and an increase in plant cycling. The commenter stated that the data also indicated that lb CO2/MWh emission rates were lower during time periods of higher generation and reduced plant cycling. The commenter stated that as utilities seek to increase reliance on intermittent renewable resources, modest flexibility in the emission limit will improve their confidence to expand renewable generation.

Commenter 0226 argued that the proposed standards for modified and reconstructed natural gas turbines are not adequately supported and, therefore, should not be applied. The commenter stated that the proposed emission standards for modified and reconstructed natural gas-based stationary CTs are based on the emission rate EPA asserts is achieved by efficient new natural gas combined-cycle units, however there is no evidence offered in the proposal that new units' performance would not more easily meet the limits simply because everything is new. The commenter stated that the potential achievability, costs and energy impacts of the standards may differ dramatically for modified and reconstructed sources versus new sources, and the proposal does not indicate that there was consideration of these and other distinguishing factors while proposing the same emission standards for both new sources and modified and reconstructed sources.
Commenter 0213 urged EPA not to allow modified stationary combustion turbines to "elect, as an alternative to the otherwise applicable numerical standard, to instead meet a unit-specific emission standard that is determined by the CAA section 111(d) implementing authority based on implementation." The commenter stated that there is no reason to create a unit-specific alternative to the proposed standard of performance for modified stationary combustion turbines. The commenter stated that the proposed standard is based on a BSER of efficient combined cycle natural gas technology. The commenter stated that CCGT technology provides baseload and load-following generation at a relatively low cost compared to other methods of fossil fuel combustion, and for many simple cycle turbines, particularly those operating at higher utilization rates, "the cost of replacement with a NGCC [CCGT] unit is likely to be cost effective based on consideration of fuel savings alone."  The commenter stated that given that CCGT is both economically reasonable and adequately demonstrated, it would be arbitrary and capricious to adopt an emission limit for modified stationary combustion turbines that is less stringent than the emission limit achievable through the application of this technology. The commenter stated that EPA should not grant sources the option of complying with a unit-specific emission standard to be established by state authorities.
The BSER for base load units has been identified as NGCC. The BSER for non-base load units has been identified as clean fuels. Whether a unit is classified as a base load unit or a non-base load unit depends on the amount of electricity sold to the grid, which is derived from a unit's nameplate design efficiency and calculated as a percentage of potential electric output. The EPA notes that the data set used to determine the final base load standard includes units that commenced operation on or after 2000. These units represent approximately 90 percent of the existing NGCC capacity. Further, these units would have begun planning on construction years prior to that. Therefore, the final base load emission standard has been demonstrated for NGCC units using technology that is over 15 years old. Finally, 40 CFR 60.15(b)(2) provides that if an owner or operator can show that it is not "technologically and economically feasible to meet the applicable standards set forth in this part" for a given unit, then the reconstruction provisions will not apply. Therefore, the commenters' concerns regarding technical feasibility are simply not an issue.
 See Sections of IX.C.4.a, IX.C.4.b, and IX.C.4.c of the final rule preamble of for a more complete discussion of the EPA's finale rationale for the BSER. See Sections IX.D.3.b, IX.D.3.c, and IX.D.3.d of the final rule preamble for a more detailed discussion of the EPA's final regarding the achievability of the final standards.
Commenter 0193 stated that there was a significant deficiency in EPA's historical analysis, specifically the underrepresentation of larger, more efficient simple cycle units in the historical database. The commenter's first LMS100 simple-cycle gas turbine power plant entered commercial operation in July 2006, thus, it did not even exist for half of the period covered in the historical analysis and certainly did not achieve market penetration at that time. As described in more detail in the commenter's comments on EPA's January 2014 proposal, the LMS100 is the most efficient, and thereby lowest CO2 emitting, simple-cycle turbine available today. The commenter stated that since introduction in 2006, LMS100 sales have increased dramatically and are likely to continue to increase due to its ability to provide 100 MW of efficient and flexible power, plus its rapid start capability that allows it to achieve full power within 10 minutes. The commenter stated that in 2013 there was an 82 percent increase in operating units (28 to 51). The commenter stated that in Southern California alone, 19 LMS 100s, came online bringing the total LMS100 count in California to 23, with an additional 5 units expected to enter the power generation market by the end of 2015. The commenter stated that these 23 units are capable of ramping up 1,200 MW/minute, while offering 2,400 MW of generation.
 The EPA specifically reviewed the operating data of LMS100 simple cycle combustion turbines. While these units tend to run more than traditional simple cycle units, the average percentage of electric sales (approximately 10 percent) was still well within the non-base load subcategory included in this final rule. In fact, no LMS100 in California has sold more than 25 percent of its potential electric output. Based on this, we have concluded that the non-base load subcategory includes sufficient flexibility for the continued use of high efficiency simple cycle turbines.
Commenter 0176 stated that NGSC units offer significant advantages over NGCC units from a water conservation perspective and for that reason should remain as a resource for intermediate loads. The commenter stated that NGCC technology is a water-intensive process that can consume hundreds of millions gallons of fresh water per year. The commenter stated that a NGCC plant can easily consume more than 270 gallons per MWh of cooling water on an annual basis, whereas a NGSC unit consumes only 41 gallons per MWh. See Wendy Wilson, Travis Leipzig and Bevan Griffiths-Sattenspiel. Burning our Rivers: The Water Footprint of Electricity, River Network (April 2012): Portland, OR. The commenter stated that while it is possible to reduce the water requirements of a NGCC by using dry cooling methods, such as an air cooled condenser (ACC) instead of an evaporative cooling tower, the use of ACC technology can result in significant losses in efficiency and net output. The commenter cited a study conducted by EPRI, that "dry cooling imposes a heat rate and lost-capacity penalty on a plant that can range up to 25% during the hottest hour of the year and exceed 8% for over 1,000 hours at a hot, arid site. On an annual basis, plant output is reduced by about 2%." See John S. Maulbetsch and Kent D. Zammit, Cost/Performance Comparisons of alternative Cooling Systems CEC/EPRI Advanced Cooling Strategies/Technologies Conference (June 1-2, 2005). 
The commenter stated that the ACC equipped plant will burn more fuel and increase air emission to produce the same net power produced by a NGCC unit using a traditional evaporative cooling tower and as demand for energy typically peaks during hot temperatures, this loss of efficiency would be significant. The commenter stated that "depending on site meteorology, the capital cost of the dry cooling system ranges from $21-$26 million for the combined-cycle plant compared to $5.7-$6.5 million for wet cooling." 
According to commenter 0176, incentivizing NGCC technology that is water intensive is a particularly important consideration in water-starved areas such as Texas. The commenter stated that In parts of the country with arid climates that are particularly susceptible to drought conditions, excessive reliance on NGCC technology may not be feasible or a desirable option. The commenter stated that Texas is experiencing its third year of intense drought conditions, with the Texas Panhandle being hit especially hard.
The commenter is mistaken that all combined cycle units consume more water than simple cycle units. Combined cycle units that use dry cooling technologies do no consume any more water than a comparable simple cycle unit. In fact, it is possible that a combined cycle unit uses less water than a simple cycle unit. Simple cycle units often use some type of water injection in the turbine itself to enhance performance and high efficiency intercooled simple cycle units have a cooling tower for interstage cooling. The EPA specifically review the emission rate for NGCC units with dry cooling technologies. The EPA was not able to identify any impact on performance since these particular units performed just as well as comparable units with wet cooling systems. Water constraints are a site specific consideration, and the final base load emission standard has been demonstrated as achievable for units using dry cooling systems.
 See sections of IX.C.4.a, IX.C.4.b, and IX.C.4.c of the final rule preamble of the EGU GHG NSPS for EPA's response to the comments regarding the rationale for BSER.
 See sections IX.D.3.b, IX.D.3.c, and IX.D.3.d of the final rule preamble of the EGU GHG NSPS for EPA's response to comments regarding rationale and achievability of the final standards.
Costs
Multiple commenters (0149, 0157, 0226, and 0242) rejected EPA's cost argument concerning NGCC technology. Commenters 0149 and 0226 stated that EPA had provided no data or analysis to support its assertions that NGCC technology is "likely to be cost effective," "likely to be made to return the unit to close to its original operating performance," and "pays for itself in fuel savings alone" for all modified and reconstructed CTs---including both NGCC and simple-cycle CTs. 
Commenter 0157 stated that EPA erred in claiming that NGCC is one of the lowest cost forms of baseload and intermediate load generation. The commenter stated that there are few if any baseload NGCC units. The commenter stated that many units operate at high capacity factors, but baseload generation is that which is running nearly 100% of the time, and few NGCC units meet that criterion. The commenter stated that the low cost of NGCC is a recent phenomenon and whether it persists for many years into the future remains to be seen. The commenter stated that EPA ignores the historic cost of gas, its price volatility, the fact that supplies are limited by current infrastructure and that many use natural gas for winter heating, making gas supplies for electric generation unreliable enough that many facilities are dual-fueled and operate on fuel oil during very cold peaking periods.
The commenter is mistaken that the EPA did not provide a cost analysis. Cost models developed by the EIA were used to compare the relative LCOE costs of simple cycle and combined cycle units at different percentage electric sales. If natural gas prices were to increase, NGCC units would become even more cost-effective relative to simple cycle combustion turbines. The multi-fuel-fired subcategory provides adequate flexibility for dual fuel units that burn 10 percent or more non-natural gas fuels.
Commenter 0131 believes EPA needs to expand its definition concerning relative cost. The commenter stated that several times EPA refers to the limitation as being 50% of the cost of a "comparable entirely new facility", however EPA avoids defining what a comparable plant is--this could be 50% of a comparable new plant using the same technology or 50% of a new plant with comparable power production. The commenter stated that the key point for clarification for coal plants is whether EPA means 50% of a new coal plant or 50% of a new NGCC with comparable power output. The commenter stated that these are very different economic criteria, and EPA needs to define its actual intent and not bury a suggestion of the definition in the section describing reasoning as to why no units will actually seek to be reconstructed.
This final rule does not amend the reconstruction provisions contained in the general provisions or establish subpart specific reconstruction provisions.
Determination of the Level of the Standard
Multiple commenters (0213, 0216, and 0282) argued that EPA's proposed standards for modified and reconstructed stationary combustion turbines are too lenient. Commenter 0213 noted that EPA's proposed performance standards for these modified/reconstructed units are the same as its proposal for entirely new units: smaller units (i.e., those with a maximum heat input equal to or less than 850 MMBtu/h) would be required to meet a gross-output emission limit of 1,100 lbs CO2/MWh, while larger units (i.e., those with a maximum heat input greater than 850 MMBtu/h) are limited to 1,000 lbs CO2/MWh. Commenter 0213 stated that like the proposed new source rule, the modified/reconstructed rule for stationary combustion turbines reflects a determination that efficient CCGT technology is the BSER for stationary combustion turbines. Commenter 0213 agreed with EPA's BSER determination for this category of EGUs, and, in principle, supported the Agency's decision not to distinguish between new, modified, and reconstructed units for the purposes of regulating CO2 emissions for these sources. Commenter 0213 cited the preamble, that efficient CCGT design is widely available and in regular use throughout the electricity sector. Commenter 0213 stated that units that modify or reconstruct should not be allowed to emit at rates greater than those achieved by the most-efficient CCGT technologies available. 
However, commenter 0213 argued that the numerical limits the Agency selected--1,100 lbs CO2/MWh for smaller units and 1,000 lbs CO2/MWh for larger units on a gross output basis--are far too lenient and do not represent the true performance capabilities of the most efficient CCGT units currently available and operating. Commenter 0213 referenced comments they submitted on the new source standard providing ample documentation regarding emission rates of the existing CCGT fleet as well as different design options currently available. Commenter 0213 stated that a full 94 percent of the CCGT units in existence as of 2011 already satisfy the emission limits EPA has selected for its modified/reconstructed rule. Commenter 0213 stated that to comport with section 111's technology-forcing mandate, and reduce CO2 emissions from modified and reconstructed stationary combustion turbines as the CAA requires, the Agency must propose limits that reflect the true performance capabilities of the most efficient CCGT models on the market.
For the reasons stated in the comment above, commenter 0216 agreed with EPA's determination that NGCC technology constitutes the BSER for modified and reconstructed natural gas-fired stationary combustion turbines, and with EPA's decision not to distinguish between new, modified, and reconstructed units in establishing emission standards for this source category. However, commenter 0216 believes that the numerical limits EPA selected in establishing this emission standard are not sufficiently stringent because they do not represent the true performance capability of the most efficient NGCC units.
Commenter 0282 stated that EPA was correct in setting the standard for modified and reconstructed sources at the same level as the proposed 111(b) standard. Commenter 0282 cited comments submitted on the section 111(b) proposal that the standard itself is indefensibly lenient and must be significantly strengthened. Commenter 0282 supported this by citing the fact that EPA has indeed selected it for existing sources that modify or reconstruct but has concluded that "the proposed standards will result in minimal emission reductions, costs, or quantified benefits by 2025." Commenter 0282 stated that it is unacceptable that the standard EPA proposed for existing modified natural gas-fired plants is lower than the section 111(d) standard these plants will have to meet if they do not modify.
Commenter 0282 incorporated their section 111(b) comments on the insufficiency of the section 111(b) natural gas-fired power plants in full, and briefly summarized some of their main points. Commenter 0282 stated that by EPA's own account, that proposed new source performance standard does not have any discernible effect on greenhouse gas emissions from the power sector and will do nothing that EPA's economic projections anticipate will not happen anyway; as such, they do not fulfill the Clean Air Act's mandate that emission standards actually reduce emissions. Commenter 0282 stated that the proposed standard is so lenient that some 98 percent of existing sources already meet it, yet EPA failed to explain why technology that is both adequately demonstrated and could achieve greater reductions, such as advanced NGCC or even more efficient NGCC technology, has not been selected. Commenter 0282 stated that data in EPA's docket demonstrate that a far more stringent emission standard is readily achievable at many existing facilities. Commenter 0282 stated that without adequate explanation, EPA decided against requiring partial (let alone anything more than 25%) CCS for natural gas plants, even though it found the technology to be adequately demonstrated for boilers and IGCCs. Commenter 0282 stated that, in creating standards that favor reliance on natural gas plants for the foreseeable future, EPA failed to consider their environmental and non-air consequences, including the greenhouse gas and adverse health effects of hydraulic fracturing and other nonconventional extraction techniques. Commenter 0282 stated that the proposal, in incentivizing increased natural gas extraction, processing, transport and combustion, undercuts the only sustainable way of producing energy, which is through renewable energy sources.
As described elsewhere in this final rule, the EPA has concluded that the final emission standard for base load units is appropriate considering the universe of units that will be subject to the requirements of this rulemaking. Specifically, the base load subcategory includes both fast-start NGCC units as well NGCC units designed primarily for steady state operation. 
Based on commercially available turbines, commenter 0193 stated that the correct cut-point between large and small gas turbines for new and modified and reconstructed units should be raised to 1,500 MMBtu/h (HHV). As explained in their attached comments to the January 2014 proposal for new units, the 850 MMBtu/h (HHV) cut-point was based on NOx emissions not CO2. The commenter stated that there are significant differences between these two pollutants and the methods and approaches to controlling their emissions. The commenter stated that gas turbine NOx emissions are a function of the combustion technology, which is dictated by the combustor design while CO2 emissions expressed in terms of lb CO2/MWh are a direct function of the overall turbine or plant efficiency. The commenter stated that because the design parameters that influence NOx formation fundamentally differ from those that affect CO2 formation, the proposed 850 MMBtu/h cut-point is inappropriate for CO2, and any "consistency" benefits of having the same cut-point as subpart KKKK are clearly outweighed by the overriding statutory requirement that EPA establish standards that are achievable and account for all source types and operations anticipated for the fleet. 
As discussed in the commenter's prior comments, an examination of the performance of all commercially available 60 Hz gas turbines (with a heat input of 250 MMBtu/h (HHV) or greater) contained in Gas Turbine World 2013 (GTW) shows that the 850 MMBtu/h (HHV) cut-point does not provide a logical break point between large and small turbines. The commenter reproduced a graph included in their comments on the January 2014 proposal depicting the estimated CO2 emissions (lb/MWh) of the fleet of commercially available gas turbines, when installed in a combined cycle configuration. The commenter stated that the vertical lines on the graph show that the 850 MMBtu/h (HHV) cut-point does not logically separate higher emitting and lower emitting turbines.
Based on existing data, the commenter stated that a more logical cut-point between large and smaller gas turbines is substantially higher. The commenter stated that, as noted by EPA, the data show that there are no combustion turbines offered for sale with turbine engine base load ratings between 1,300 to 1,800 MMBtu/h. The commenter stated that this range presents a clear data-driven size and performance basis for distinguishing between large and small gas turbines, and that a dividing line that falls between these heat input values of 1,300 MMBtu/h (HHV) and 1,800 MMBtu/h (HHV) would appropriately divide turbines based on their size and emissions performance capabilities.

Commenter 0212 believes that EPA should establish a single performance standard for all stationary gas turbines, regardless of turbine size or rated heat input. Commenter 0212 stated that separate standards for small and large turbines undermine the incentive for technology innovation and discourage investment in more efficient technologies. As an alternative to increasing the standards of performance for large and small natural gas turbines, commenter 0192 supported a single performance standard of 1,200 lb CO2/MWh for all modified and reconstructed natural gas turbines

Commenter 0213 argued that EPA should not create a separate emission limit for smaller CCGTs (i.e., those with a maximum heat input under 850 MMBtu/hr).The commenter stated that if such as standard is created it should not exceed 1,000 lbs CO2/MWh. The commenter stated that under this scenario, EPA must also provide that multiple smaller units at the same physical site will be considered a single source for the purpose of calculating emissions.
For the reasons stated in the preamble to this final rule, we are not finalizing size-based subcategories for combustion turbines. The EPA's rationale for establishing a single size category include that the EPA was not able to identify a reasonable cut-point, that any cut-point has the potential to distort the market, and that the final emission standard is achievable by both small and large NGCC units.
Commenter 0149 noted that, in the proposal, EPA solicited comment on whether to establish a variable standard that would apply to different operating modes. The commenter stated that a variable standard may make sense in theory, but raises many questions that EPA has not addressed. The commenter stated that it is difficult to assess a variable standard without reference to the actual standard, which EPA has not proposed in the context of this rulemaking, and further, EPA has not proposed a time scale over which the different operating modes would be measured. The commenter stated that unless and until EPA provides more details about a variable standard for new NGCC units--and provides stakeholders sufficient notice and an opportunity to comment--the Agency cannot finalize a variable standard for these units.
For the reasons discussed in the preamble to this final rule, the EPA is finalizing a two-tiered emission standard for non-base load and base load units. Based on additional analysis and comments, the base load standard is only applicable when the percentage electric sales threshold is exceeded on both an annual and 3-year average. 
7.2.1 Large Units Limit of 1,000 lb CO2/MWh; solicited comment on range of 950-1,100 lb CO2/MWh
Commenter 0215 stated that based on its analyses of the 2007 to 2013 data about 10 percent of NGCC units that were designed and built with truly modern technology cannot meet a 1,000 lb CO2/MWh-gross limit and that only 96 percent of large (more than 850 MMBtu/h) units can meet a 1,100 lb CO2/MWh-gross limit (See: Smith Turbine Report at 2 & 7, Fig. 1; Attachment A at 3]. The commenter stated that these results are substantially below the 99 percent achievability target that EPA has historically applied in NSPS rulemakings and would require a significant number of NGCC units to retire if they became subject to the Proposed Standards. The commenter stated that in order for 99 percent of turbines with truly modern NGCC technology to comply, EPA would need to finalize an NSPS of over 1,400 lb CO2/MWh, and although applying a one-third sales criterion to exclude certain turbines did reduce the standard that would be required for 99 percent achievability, (see Smith Turbine Report) the proposed 1,000 lb CO2/MWh standard was still not achievable for new NGCC units even with a one-third sales criterion, thus many more units that are older, which do not have high temperature materials, will be unable to meet the proposed limit.
The EPA notes that the commenter is mistaken about how the EPA has established emission standards for coal-fired EGUs. In recent NSPS amendments, the EPA first identified the best performing units (i.e., the units with the lowest emission rate) and then determined the 99 percent confidence level strictly based on the emission rate data from the best performing unit(s). 
Commenter 0215 stated that EPA should analyze emissions from NGCC units at least as far back as 1990, explain what such units can realistically accomplish through retrofitting newer technology with higher temperature materials (including citing specific examples of instances where such work was undertaken), and then re-propose a limit so that the commenter can make informed comments on EPA's justification. The commenter stated that design efficiencies of NGCC units built before 2000 would need to improve their efficiencies from 53 to 55 percent to as much as 60 percent to meet the proposed NSPS, but these units have no available means to improve heat rate back to their out-of-the-box design levels, let alone to improve their operating efficiencies by 5 to 7 percentage points through overhauls or other refurbishments. 39487
While the commenters' points may e applicable to modified units, the cost threshold for reconstructed units would allow the owner or operator to upgrade components sufficiently to achieve the final emission rate. 
7.2.2 Initial Performance Test in Addition to Ongoing Average
Rather than establishing separate performance standards for large and small turbines and relying solely on a 12-month average emissions standard, commenter 0212 recommended a hybrid approach consistent with the approach for which EPA asked for comment. The commenter recommended combining (1) an initial performance test with (2) a 12-month average performance standard. The commenter recommended that within 180 days of initial firing following a reconstruction, a combustion turbine facility would demonstrate compliance with an initial performance standard while operating at full load, ISO corrected conditions. 
The commenter recommended that after demonstrating compliance with the initial performance standard, the facility would be subject to an ongoing 12-month average emission standard, beginning 365 days after initial firing, like the ongoing performance standard proposed by EPA. The commenter stated that the ongoing performance standard would be less stringent than the initial performance standard to allow for the diminished efficiency and higher emissions associated with increased cycling, part load operation, oil combustion, normal plant degradation, as well as variations in ambient temperature and other environmental factors beyond the control of the plant operator.
The commenter stated that there are several advantages of a hybrid approach. The commenter recommended an initial performance standard to ensure that companies select high efficiency generating technologies when making their initial construction or reconstruction decisions. The commenter stated that if the initial performance standard for new turbine facilities is set at a level of 900-950 lb/MWh, companies can design to this specification and have a high degree of confidence that the plant will meet this performance standard at baseload conditions. The commenter stated that correcting to ISO conditions would ensure that a plant is not disadvantaged by virtue of its high elevation or ambient temperature. The commenter stated that duct firing should be excluded during the initial performance test for a standard of 900-950 lb/MWh to be reasonably achievable. The commenter stated that EPA should recognize that an existing modified or reconstructed facility may not be able to achieve the same level of performance as a new, state-of-the-art combined cycle facility.
The commenter stated that an ongoing performance standard of 1,100 lb/MWh (with reasonable accommodations for alternative fuel use and startup and shutdown) would ensure that a new turbine facility plant is operated efficiently and well maintained, while providing a reasonable compliance margin if market conditions change and the facility needs to cycle more. The commenter expected that EPA would hear concerns from a number of stakeholders regarding the potential effects of degradation and cycling on a plant's CO2 emission rate (The California Energy Commission (CEC), for example, assumes a 4% heat rate degradation factor for a natural gas combined cycle facility over 20 years of operation. See "Comparative Costs of California Central Station Electricity Generation," August 2009. The commenter stated that the emission rate of a NGCC facility will increase with (1) part load operation, (2) the use of higher carbon content backup fuels, (3) startup and shut events, (4) degradation, and (5) other factors like higher elevation and ambient temperatures. With the approach the commenter suggested, they expected a new plant to operate well below the ongoing performance standard because the economics of the power sector encourage efficiency and the initial performance standard will require companies to invest in high efficiency, combined cycle technologies. The commenter stated that in setting an ongoing performance standard for modified and reconstructed facilities, EPA should recognize that an existing modified or reconstructed facility may not be able to achieve the same level of performance as a new, state-of-the-art combined cycle facility.
The commenter noted that EPA, in its January proposal, suggested that smaller capacity NGCC facilities may run in simple cycle mode for short periods of time, increasing their average CO2 emission rate. The commenter stated that this was, in part, the justification for proposing a higher performance standard. The commenter stated that a hybrid approach would address this concern. The commenter stated that the initial performance test would be conducted in combined cycle mode, while the on-going performance test would accommodate simple cycle operation for short periods of time.
According to the commenter, reducing the carbon intensity of the electric sector presents a unique challenge, which is different from past regulatory efforts aimed at conventional air pollutants. The commenter stated that the daily emissions performance of the fleet matters less than the industry's performance over many years or even decades, and it is in this context that the commenter recommended a hybrid approach with an initial performance standard that will ensure that the right investment decisions are made at the outset. The commenter stated that EPA asked for comment on this same approach in its proposed rule for new stationary gas turbines: "[w]e are considering and requesting comment on requiring an initial performance test for stationary combustion turbines in addition to the 12-operating month rolling average standard". The commenter stated that requiring an initial compliance test that is numerically more stringent than the annual standard for new combined cycle facilities would insure that the most efficient stationary combustion turbines are installed, and a less stringent 12-month rolling average standard would be set at a level that would take into account actual operating conditions.
The commenter noted that EPA indicated that it received feedback on its January 2014 proposal suggesting that the emissions data that it used in developing the natural gas-fired stationary combustion turbine standards did not completely account for degradation in performance over the entire life of an NGCC. Also, the commenter noted that NGCC units are expected to operate differently in the future due to the increased percentage of power generated from renewable sources, such as wind and solar, and an initial performance test combined with a less stringent on-going performance standard would 
Commenter 0232 was concerned that EPA's proposed standards do not include an initial performance demonstration to ensure that modified and reconstructed EGUs use the most efficient and least-polluting generating technologies available. The commenter stated that the proposed standards require only that these EGUs meet the applicable average emission standard after the first 12-month compliance period has ended. The commenter stated that for natural gas combustion turbines, these average standards do not even reflect the performance of the most efficient NGCC facilities currently being operated, let alone the best performance of a NGCC facility operating under optimal conditions. The commenter stated that in order to ensure that modified and reconstructed EGUs incorporate the most efficient generating technologies available, it is essential that EPA augment the average annual performance standards with an initial performance test--consistent with their comments on EPA's proposed standards for new EGUs. The commenter stated that this is precisely the approach that state permitting authorities and EPA have undertaken in recent PSD permits for NGCC facilities, which require that the permitted facilities meet a stringent initial CO2 performance standard within 180 days after startup (as currently required for other pollutants in the General Provisions of the NSPS. 
The commenter stated that, because the purpose of these initial performance tests is to ensure that the facilities are using the lowest-emitting equipment and processes available and therefore perform as efficiently as possible under ideal operating conditions, the tests are conducted while the facility is operating at 90 to 100% of rated capacity and are normalized for temperature, pressure, and other variables. The commenter stated that there are circumstances, such as in the case of "peaking" power plants, where normal operation results in variable and on average higher emission rates than the plant can achieve under optimal operating conditions. The commenter stated that a continuously applicable emission standard for such units would be set at a level that reflected this variability. The commenter stated that the incorporation of an initial performance test that reflects the emission rate achievable using the best system of emission reduction when a plant is operating at optimal conditions ensures that facilities are built, reconstructed, or modified using the lowest-emitting technologies and operating systems available, fulfilling the technology-forcing and pollution-minimizing purposes of Section 111. The commenter stated that they are an essential component of the carbon pollution standards for newly constructed, modified, and reconstructed units, ensuring that the standards fulfill the section 111 statutory requirements and case law.
According to the commenter, requiring an initial performance test is not only reasonable for modified or reconstructed EGUs, it is also fully consistent with similar requirements for other pollutants regulated under the NSPS. The commenter stated that subpart KKKK currently requires that natural gas combustion turbines complete initial performance tests no later than 180 days after startup to demonstrate compliance with emission standards for nitrogen oxides and sulfur dioxide, and that EGUs covered by subpart Da are similarly required to complete initial performance tests within 180 days of startup for particulate matter, sulfur dioxide, and nitrogen oxides. The commenter stated that EPA has not explained why it has departed from this time-tested requirement in the proposed standards, especially when PSD permits for have shown that initial performance tests are both feasible and desirable to ensure that NGCC facilities (and potentially other EGUs) incorporate the most efficient available generating technologies.
Multiple commenters (0131, 0193, and 0242) opposed the imposition of an additional annual short-term performance test. Commenter 0193 argued that the imposition of a short-term annual test, in addition to a 12-month rolling average standard, would add no significant value sufficient to offset the increased cost and uncertainty created by the test. Commenter 0192 stated that New NGCC units can readily meet EPA's proposed emission standards for new units when operating at ideal conditions, specifically when first installed in a new and clean condition, operating at baseload and corrected to standardized conditions. Commenter 0193 stated that the challenge is in consideration of the 12-month rolling average that encompasses all possible operating scenarios encountered by a real world operating facility. Commenter 0193 stated that establishing a short-term annual test would impose additional regulatory burden with no associated benefit in the form of emission reductions. 
Commenter 0131 noted that the rolling average was developed specifically to provide an averaging due to conditions like part load and heavy load (optimum) operation. Commenter 0131 stated that adding an additional short test period does little to provide usable data or verification, and that owners will generally seek to make sure their machines are operated and installed using best practice. The commenter stated that EPA should recognize that best practice is necessary to optimize performance both operationally and economically--the latter driver is the most important one.
Commenter 0242 also disagreed with the need for a short term compliance test. Commenter 0242 was not sure whether EPA was suggesting that a more stringent standard than those proposed could later be set based on a short term compliance test, or whether EPA was proposing to set a short term limit based on the compliance test. Commenter 0242 questioned whether EPA considered that these units will not always be able to operate at their most efficient short term rate. Commenter 0242 stated that If EPA continued forward with this proposal, and so few sources would be affected as EPA suggested, then states should be able to set the limitation at a level that would account for different operating conditions; it should be the state's responsibility to set those limits based on each unit's unique circumstances. Commenter 0242 was not in agreement with short-term requirements being imposed under this proposal. Commenter 0242 stated that the proposal addresses CO2 emissions for which there is no basis or reasoning for requiring any short-term requirements to limit emissions and provides no logic as to why short-term limits in addition to long-term limits will address climate change
Commenter 0212 disagreed with EPA on establishing a separate standard for load-following (i.e., intermediate capacity factor) NGCC EGUs. The commenter stated that the more stringent standard that EPA is considering would apply only during periods of high annual capacity factors, and a less stringent standard would apply during periods of intermediate load (e.g., when electric sales are between 33 to 60 percent of the potential electric output). The commenter believes that an initial performance test combined with an on-going performance standard (hybrid approach) would be far simpler to implement and enforce than this approach. The commenter believes this issue highlights one of the key advantages of a hybrid approach. The commenter stated that there is far less uncertainty regarding the performance of a new NGCC facility operating at full load under ISO conditions versus an older facility operating at variable load conditions. The commenter stated that in some cases, fast start and hot restart capability can result in a slightly lower base load efficiency; however, the effect is modest and would not prevent a facility designed for base-load and load-following operation from achieving a standard between 900-950 lb/MWh. The commenter stated that by establishing an initial performance standard (consistent with EPA's BSER determination), EPA will ensure that companies invest in the high-efficiency generating technologies, which in turn, reduces the burden on the on-going performance standard to accommodate different levels of cycling and intermediate load conditions. The commenter stated that the on-going performance standard can be set at a level (like 1,100 lb/MWh) that is reasonably achievable under a wide range of operating conditions, while still relying on the initial performance test to drive the best technology choice. The commenter stated that the alternative approach that EPA is suggesting, with separate standards for different operating conditions, could have the unintended consequence of encouraging less efficient technologies and less efficient operation.
While in concept an initial performance test in conjunction with an on-going continuous compliance standard appears appealing, the EPA disagrees that is it appropriate for the final rule. The universe of NGCC units covered by the base load subcategory includes fast-start NGCC units and NGCC units designed for steady state base load operation. An initial performance test based on the performance of the latter would preclude the use of fast-start NGCC units even though, in their intended role of frequent starting, stopping, and cycling, their actual performance may be superior to a NGCC unit designed for steady state operation.
7.2.3 Load Following/Intermediate vs. Base Load Standard
Commenters 0187 and 0203 supported EPA's proposed approach to establish separate standards for load-following NGCC EGUs.   Commenter 0187 also believes that modified combustion turbines and modified boilers fall into this category.  Commenters 0187 and 0203 stated that this approach would apply a more stringent standard during periods of high capacity factors and a less stringent standard during periods of intermediate capacity factors (e.g., 33 to 60%). Commenter 0187 noted that many boilers that used to be base-loaded are now load-following at times. Commenter 0187 stated that load factor has a significant impact on unit heat rate and the efficiency of units will be less than in the baseline period when they were base-loaded.
Commenter 0193 stated that EPA was correct to examine whether a separate standard should be established for load-following (i.e., intermediate capacity factor) NGCC EGUs, but was concerned about its implementation and enforcement. The commenter stated that the more stringent standard would apply only during periods of high annual capacity factors and a less stringent standard would apply during periods of intermediate load (e.g., when electric sales are between 33 to 60 percent of the potential electric output). The commenter stated that this approach addresses two potential issues with the standards in the January 2014 proposal:
(1)  certain NGCC units are designed to be highly efficient when operated as load-following units, but these design characteristics reduce the efficiency at base load. Conversely, the NGCC units with the highest base load design efficiencies are not necessarily as efficient as NGCC designed and intended to be used as load-following EGUs. Therefore, a full-load efficiency performance test would not necessarily result in the lowest CO2 emissions in practice. 
(2) NGCC units operating as load-following EGUs are inherently less efficient than NGCC units operating at base load. Establishing a standard that varies with load would assure that NGCC units that are operated as base load units are as efficient as possible and still account for inherent lower efficiencies at part-load conditions. 
The commenter noted that gas turbine installations that operate at base-load would generally demonstrate a higher 12-month average efficiency (i.e. lower CO2/MWh emission rate) than load-following installations that operate between 33 and 60 percent capacity factor on average when considering the broader range of operational conditions the intermediate/load following plant would experience during the 12-month averaging period. The commenter stated that in soliciting comments on this proposal, EPA recognized the difficulty of establishing a one-size-fits-all standard for combustion turbines that are increasingly being designed and purchased for different power needs. While the commenter supported EPA's exploration of this issue, the commenter remained concerned that a load-following standard would be difficult to implement and enforce, as annual capacity factors would likely vary significantly from year to year, and in fact would quite possibly vary from month to month as the 12-month average is recalculated on a 30-day basis. The commenter stated that this uncertainty for purchasers, operators and regulators may diminish the value of a load-following standard.

Commenter 0195 generally supported EPA's proposed approach to set a separate standard for load-following (i.e., intermediate capacity factor) NGCC units, but the commenter also had some concerns. The commenter cited the preamble discussion in the proposal, that a more stringent standard would apply only during periods of high annual capacity factors and a less stringent limit would apply during periods of intermediate load. The commenter stated that EPA explained that this approach would address the fact that when NGCC units are designed to be load-following units, these design characteristics will impede efficient operation at base load and vice versa, and that NGCC units operating as load-following EGUs are not as efficient as base load NGCC units and thus approach would address this situation. On a conceptual level, the commenter stated that a flexible standard that accounts for varying efficiencies due to how the units are operated is appropriate and reasonable. The commenter felt there was not sufficient specificity in the current proposal on how a flexible standard would work in practice for load-following units. The commenter stated that EPA needs to conduct a supplemental rulemaking to provide more detail on this type of standard so that industry can provide meaningful input.
Commenters 0213 and 0232 noted that EPA crafted the modified/reconstructed rule proposal so as to exclude units that operate in peaking mode and argued there is no basis in law or policy to exempt units from regulation simply because they operate less frequently (and hence less efficiently). The commenters stated that section 111(b) expressly permits EPA to create different emission limits for different kinds of units within a given regulatory category ("The Administrator may distinguish among classes, types, and sizes within categories of new sources for the purpose of establishing such standards"). The commenters stated that rather than simply exempting units from regulation on the basis of how frequently they run over the course of a calendar year or on a three-year rolling average basis, the Agency should cover all sources that supply (or are designed to supply) any amount of electricity to the grid, and should distinguish among these sources on a functional basis. The commenters, therefore, proposed the same three-tiered regulatory model for modified and reconstructed stationary combustion turbines that they discussed in their comments on the new source standard, which recommends that EPA recognize three subcategories of NGCC facilities--baseload, load-following/intermediate, and peaking EGUs--and establish the following separate performance standards for each: 
 Peaking units (defined as affected EGUs that operate less than 1200 hours per year) would be subject to a net output-based emission limit of 1,100 lb CO2/MWh.
 Intermediate/load-following units (defined as EGUs that operate between 1,200 and 4,000 hours annually) would be subject to a net output-based emission limit of 875 lb CO2/MWh.
 Baseload units (defined as EGUs that operate over 4,000 hours annually) would be subject to a net output-based emission limit of 825 lb CO2/MWh.
The commenters stated that the recommended standards are based on the reported performance of NGCC units in each of these subcategories and are entirely achievable and well within a reasonable cost range given the technology that is currently available, and modified and reconstructed units should be required to meet them. The commenters referred to their comments on the proposed carbon pollution standards for new EGUs. See: EPA-HQ-OAR-2013-0495-9514.
Commenter 0215 stated that EPA has acknowledged that CO2 emissions increase at lower loads, so in principle this might diminish the number of units that cannot meet the Agency's proposed NSPS for both reconstructed and modified NGCC units. The commenter stated that EPA adopted a similar approach, for similar reasons, in its NSPS for NOx for Subpart KKKK units, setting the standard for operations at peak load (100 percent of the manufacturer's design capacity of the combustion turbine at ISO conditions). The commenter stated that EPA's suggested approach (as well as the similar approach that EPA used in establishing the NSPS for NOx for NGCC Subpart KKKK units) results in "little advantage..." [See Smith Turbine Report at 5].
The final applicability criteria included in the final rule include all new and reconstructed combustion turbines regardless of actual electric sales. The EPA has finalized standards accommodating base load and non-base load operation. The BSER for base load units has been identified as NGCC. The BSER for non-base load units has been identified as clean fuels. Whether or not a unit is one category or the other is based on the amount of electricity sold to the grid, derived from a unit's nameplate design efficiency calculated as a percentage of potential electric output.
Commenters 0142 and 0242 suggested that if the EPA establishes separate standards for load-following NGCC units, the EPA should also establish separate standards for coal-fired EGUs based upon how they are dispatched. Both commenters noted that for gas turbines, the EPA is soliciting comment on whether a separate standard should be established for load-following (i.e. intermediate capacity factor) NGCC EGUs, where the more stringent standard would apply only during periods of high annual capacity factors and a less stringent standard would apply during intermediate load periods (e.g., when electric sales are between 33 to 60% of potential electric output). The commenters stated that this would address the fact that certain NGCC are designed to be highly efficient when operated as load-following units, but due to the design characteristics efficiency is reduced at base load (and vice-versa). The commenters stated that NGCC units operating as load-following are inherently less efficient than base load units. The commenters agreed that EPA is correct that a unit's dispatch will affect its emissions but argued that these concepts are true for most fossil-fuel fired generators, not just NGCC units. Commenter 0242 found it difficult to rationalize why such considerations would potentially be given to natural gas units and not coal-fired boilers. Commenter 0242 found it particularly perplexing when they understood that some coal-fired units may have to trigger the "modify" or "reconstruct" definition to achieve compliance with section 111(d) requirements.
The EPA disagrees that tiered emission standards are appropriate for coal-fired EGUs. Coal-fired units, unlike natural gas-fired units, have the option to burn fuels with lower CO2 emission rates. The use of natural gas or distillate oil would reduce the emission rate of a coal-fired EGU by approximately 40 to 20 percent respectively. The use of either of these fuels would more than offset any increase in the emission rate that results from operating at part-load conditions. A further difference is that stationary combustion turbines include two distinct technology choices, simple cycle and NGCC technology, with different emission profiles and operating characteristics. This was a factor in determining a separate BSER for non-base load and base load units.  
