Effects of Converting Nuclear Plants in Midwest to Synchronous Condensers

EFFECTS OF CONVERTING NUCLEAR PLANTS IN MIDWEST TO SYNCHRONOUS CONDENSERS

A THESIS SUBMITTED TO THE FACULTY

OF THE GRADUATE SCHOOL

OF THE UNIVERSITY OF MINNESOTA BY

TAHNEE MILLER

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE OF MASTER OF SCIENCE

Paul Imbertson, Advisor

Ned Mohan, Advisor

June 2018

Acknowledgements I’d like to express my appreciation to the following individuals for their support on the project:

• Steve Seitz at MISO for diligently following up to help get access to the models and other information necessary to complete the research. • Douglas Brown at Siemens PTI for his time helping with the dynamic analysis and his support of the project. • Chad Snyder at GE Power for his knowledge of synchronous condensers and his willingness to share some of his experiences of past projects. • Ned Mohan at the University of Minnesota for his ideas and patience. • My husband, Greg, for his endless support and encouragement and for always going over and above to help me find the time and mental capacity to work on my thesis. The kids never noticed I was too busy to play with them when I was working, because they were having too much fun with you! Thank you for always treating my goals with the same rigor as if they were your own.

Abstract Conventional power plants are being retired and replaced with power generation from new renewable generation facilities around the country. One significant drawback to this trend is the loss of system inertia, which conventional power plants are typically able to provide better than wind and solar plants, and the corresponding decrease in grid frequency stability. One way to provide system inertia is with a synchronous condenser. This thesis investigates the potential for converting five nuclear plants in the Midwest into synchronous condensers upon retirement. A transmission system software model was used to simulate disturbances in frequency with the electric grid in the existing system configuration, with the studied facility retired, and with the studied facility converted to a synchronous condenser. Synchronous condenser conversion was determined to have the greatest positive impact on local frequency stability in areas that are close to a large amount of load and renewable generation projects.

Table of Contents Acknowledgements ...... i

Abstract ...... ii

List of Tables ...... v

List of Figures ...... vi

1 Introduction ...... 1

2 Background ...... 2

2.1 Synchronous Condensers ...... 2

2.2 Frequency Stability ...... 3

2.3 Synchronous Condenser Conversion ...... 3

2.4 MISO Planning Area ...... 4

3 Analysis ...... 5

3.1 Existing MISO Planning Model ...... 5

3.2 Selection of Study Cases ...... 6

3.2.1 Clinton Generating Facility ...... 7

3.2.2 Duane Arnold Generating Facility ...... 7

3.2.3 Palisades Generating Facility ...... 7

3.2.4 Prairie Island Generating Facility ...... 8

3.2.5 Kewaunee Generating Facility ...... 8

3.3 Study Procedure ...... 10

3.3.1 Create Modified Models ...... 10

3.3.2 Simulate Frequency Disturbances ...... 12

3.3.3 Measured Quantities ...... 13

3.3.4 Analysis Limitations ...... 14

4 Results ...... 15

4.1 Clinton Results ...... 15

4.2 Duane Arnold Results ...... 20

iii

4.3 Palisades Results ...... 25

4.4 Prairie Island Results ...... 30

4.5 Kewaunee Results ...... 35

5 Conclusions ...... 40

5.1 Summary of Findings ...... 40

5.2 Conclusions Drawn By Results...... 40

5.3 Future Work ...... 41

References ...... 42

List of Tables Table 1: MTEP16 Model Base Case Summary ...... 5 Table 2: MISO Nuclear Generation Facilities ...... 6 Table 3: MISO Nuclear Generation Facilities Selected For Evaluation ...... 8 Table 4: Frequency Disturbance Simulations ...... 12 Table 5: Clinton Results Summary – Summer Shoulder Case ...... 15 Table 6: Clinton Results Summary – Summer Peak Case ...... 15 Table 7: Duane Arnold Results Summary – Summer Shoulder Case ...... 20 Table 8: Duane Arnold Results Summary – Summer Peak Case ...... 20 Table 9: Palisades Results Summary – Summer Shoulder Case...... 25 Table 10: Palisades Results Summary – Summer Peak Case ...... 25 Table 11: Prairie Island Results Summary – Summer Shoulder Case ...... 30 Table 12: Prairie Island Results Summary – Summer Peak Case ...... 30 Table 13: Kewaunee Results Summary – Summer Shoulder Case ...... 35 Table 14: Kewaunee Results Summary – Summer Peak Case ...... 35

List of Figures Figure 1: MISO Map ...... 4 Figure 2: Wind Generation With Selected Sites Overlayed ...... 9 Figure 3: Frequency Measurement Quantities ...... 13 Figure 4: Clinton - Generator Trip in Summer Shoulder Case ...... 16 Figure 5: Clinton - Generator Trip in Summer Peak Case ...... 17 Figure 6: Clinton - Load Trip in Summer Shoulder Case ...... 18 Figure 7: Clinton - Load Trip in Summer Peak Case ...... 19 Figure 8: Duane Arnold - Generator Trip in Summer Shoulder Case ...... 21 Figure 9: Duane Arnold - Generator Trip in Summer Peak Case ...... 22 Figure 10: Duane Arnold - Load Trip in Summer Shoulder Case ...... 23 Figure 11: Duane Arnold - Load Trip in Summer Peak Case ...... 24 Figure 12: Palisades - Generator Trip in Summer Shoulder Case ...... 26 Figure 13: Palisades - Generator Trip in Summer Peak Case ...... 27 Figure 14: Palisades - Load Trip in Summer Shoulder Case ...... 28 Figure 15: Palisades - Load Trip in Summer Peak Case ...... 29 Figure 16: Prairie Island - Generator Trip in Summer Shoulder Case ...... 31 Figure 17: Prairie Island - Generator Trip in Summer Peak Case ...... 32 Figure 18: Prairie Island - Load Trip in Summer Shoulder Case ...... 33 Figure 19: Prairie Island - Load Trip in Summer Peak Case ...... 34 Figure 20: Kewaunee - Generator Trip in Summer Shoulder Case ...... 36 Figure 21: Kewaunee - Generator Trip in Summer Peak Case ...... 37 Figure 22: Kewaunee - Load Trip in Summer Shoulder Case ...... 38 Figure 23: Kewaunee - Load Trip in Summer Peak Case ...... 39

1 Introduction Today’s electric grid in the United States is drastically different than the electric grid of the past. While in the 20 th century coal, nuclear, and other large-scale thermal plants were being built to generate large amounts of power, today these same plants are being replaced with other types of generation, including generation from new wind and solar facilities being constructed around the country. Many of the nuclear facilities built in the 1970s and 1980s are being shut down due to changing environmental policies, economics, maintenance concerns, or because they are at the end of their fuel cycle [1].

Shutting down nuclear facilities may lead to decreased system stability due to the loss of inertia provided by the plant. Inertia is crucial to helping the grid appropriately respond to major frequency excursions caused by events such as if a large generator were tripped offline suddenly or if there was an abrupt change in the amount of power being requested by loads, like homes and businesses. While wind and solar plants are subjected to an increasingly conservative array of frequency response guidelines [2], a large generating facility is much better at providing the system inertia needed to ride through a system disturbance causing a large frequency excursion.

One way to provide system inertia is with a synchronous condenser. A synchronous condenser is a large generator that is decoupled from the turbine, allowing it to provide a source of system inertia and reactive power, but no real power.

This thesis investigates the potential for converting nuclear plants which are recently retired or may be retired into synchronous condensers to improve power system stability and utilize the electrical equipment that is already in place at retired facilities. The research presented here focuses on the potential for implementing this solution in the transmission system operated by the Midcontinent Independent System Operator (MISO), a Regional Transmission Organization (RTO) operating electric power across all or parts of 15 states primarily in the Midwest.

2 Background The following sections describe some of the background information for the research presented.

Synchronous Condensers Synchronous condensers have been a part of the electric grid in the United States and around the world for more than a century [3]. A synchronous condenser is a dynamic device that consists of a synchronous motor that is decoupled from the prime mover. While a typical generator converts mechanical power to electric power, a synchronous condenser converts stored rotating energy to produce or adjust other conditions in the electric grid besides electric power to improve the quality of the grid. Some of the advantages that synchronous condensers provide for the bulk power system are listed below.

• Dynamic Reactive Power A synchronous condenser is primarily used as a reactive power source. Since a synchronous condenser is decoupled from the governor, no real power is produced, but reactive power is. This allows the synchronous condenser to produce either capacitive or inductive reactive power, as requested by the grid operator, by adjusting the electric field current. • Rotational Inertia Synchronous condensers can provide inertia to the power system like conventional power plants. Inertia is important for the power system’s frequency stability. The rotating energy can be delivered to the grid as real power when frequency drops and can be absorbed from the grid as real power when frequency rises to improve stability. • Steady-State and Dynamic Voltage Control Synchronous condensers can provide voltage support to the power grid during system faults or other events which cause an increase or decrease in the system voltage. The voltage support provided by the synchronous condenser can help the power system ride through the event. • Increased Short Circuit Strength A synchronous condenser increases the short circuit strength at the point of installation on the grid like a conventional power plant. A strong grid, rather than a weak one, is generally more stable and has less of a risk for transient voltage problems due to faults or other events. Increased short circuit strength can also increase the transmission capacity.

One of the prime advantages of synchronous condensers in today’s transmission grid is its ability to provide inertia. As conventional power plants close and wind and solar projects are built to meet ever-increasing requirements for renewable energy and to satisfy increases in electricity demand, the transmission grid is becoming more and more dependent on renewable generation. One service that wind and solar projects are not able to provide to the system is the rotational inertia required for a stable electric system. While wind and solar manufacturers are improving technology to provide “synthetic inertia” [4], it is not currently at a point where the loss of inertia when conventional power plants are shut down will go unnoticed and, in general, they lack the inertia that can help the grid to ride through a major fault [5].

Frequency Stability Fluctuations in the frequency of the power system are caused by changes in load or generation that occur too quickly for the system generators and their controls to respond to the fluctuation [6]. When a large increase or decrease in real power takes place, the system frequency decreases or increases as a result of the change. The characteristics of the electric grid affect the ability of the transmission system to ride-through these excursions in frequency.

Local frequency stability can be measured by looking at the frequency nadir, settling frequency, and rate of change of frequency (ROCOF). These quantities are described in more detail in the analysis section.

Synchronous Condenser Conversion There are several companies that have converted conventional power plants into synchronous condensers. One company, General Electric (GE), converts existing power plants by using a variable frequency drive (VFD) to drive the existing generator to synchronous speed. The generator is decoupled from the prime mover and controls are added to the facility. Engineering costs for this type of conversion may be in the range of 8 to 12 million dollars. Operation and maintenance costs would be low since there is no wear on the decoupled turbine. Personnel needs are typically much less for a power plant after conversion to a synchronous condenser since there is less maintenance on the facility. Converting an existing conventional power plant can also provide a return on the initial plant investment and can be an easy and inexpensive transition if the plant is well-maintained. Commonwealth Edison converted two units at the Zion Nuclear Power Station near Chicago to synchronous condensers in 1997. One consideration to this type of conversion is there is generally inadequate compensation for frequency response or inertia, making it hard to recover the cost of conversion for plant owners.

MISO Planning Area MISO is a not-for-profit organization that ensures the reliable delivery of electricity at the lowest cost in the power system in fifteen states and Manitoba [7]. MISO works with the neighboring regional transmission organizations PJM and Southwest Power Pool as well as the Southeastern Regional Transmission Planning Region to coordinate transmission planning efforts.

FIGURE 1: MISO MAP

3 Analysis The study procedure described below was used to determine the potential impact of converting nuclear facilities into synchronous condensers.

Existing MISO Planning Model A planning software model of the electrical system was requested from MISO in Power Systems Simulator for Engineers (PSS®E) format to be utilized for academic research. MISO provided the load flow and dynamic models of the MISO system used for the MISO Transmission Expansion Plan (MTEP) 2016 study [8]. The MTEP is an annual report completed by MISO providing evaluation and recommendations for various transmission expansion projects and reporting on various other specific issues relevant for each cycle of the report. The models provided included the following cases, with base characteristics as shown in Table 1.

TABLE 1: MTEP 2016 MODEL BASE CASE SUMMARY Maximum Modeled Case # of Buses Generated Generated Power (MW) Power (MW) 2021 Summer Shoulder with 40% 76,715 905,798 598,362 Wind Generation 2021 Summer Peak 77,189 922,010 726,961

Selection of Study Cases The MISO system presently contains several nuclear generation facilities, as shown in Table 2. Table 2 also shows those facilities in the MISO region that have been decommissioned.

TABLE 2: MISO NUCLEAR GENERATION FACILITIES Nuclear Generation Facility Status Utility State Elk River Decommissioned in 1968 GRE Minnesota Pathfinder Decommissioned in 1967 NSP South Dakota La Crosse Decommissioned in 1987 Dairyland Wisconsin Kewaunee Decommissioned in 2013 Dominion Wisconsin Big Rock Point Decommissioned in 1997 Consumers Michigan Fermi 1 Decommissioned in 1972 DTE Michigan Planned for Shutdown Palisades Entergy Michigan in 2022 Monticello Operating Xcel Minnesota Prairie Island 1 Operating Xcel Minnesota Prairie Island 2 Operating Xcel Minnesota Point Beach 1 Operating NextEra Wisconsin Point Beach 2 Operating NextEra Wisconsin Fermi 2 Operating DTE Michigan Clinton Operating Exelon Illinois Duane Arnold Operating NextEra Iowa Callaway Operating Ameren Missouri Arkansas Nuclear One 1 Operating Entergy Arkansas Arkansas Nuclear One 2 Operating Entergy Arkansas Grand Gulf Operating Entergy Mississippi River Bend Operating Entergy Louisiana Waterford Operating Entergy Louisiana

Publicly available information was studied for each of these nuclear facilities, with the purpose to identify those which may be candidates for conversion to a synchronous condenser. Key characteristics of a facility which may be a good candidate for this conversion process are:

• Proximity to high amounts of wind penetration • Proximity to load • Recorded issues at the plant, such as safety concerns or previously identified retirement • Status of the facility itself in the case of those plants which are already retired

The following sites described in detail below were chosen to be the most likely candidates for conversion to synchronous condensers.

3.2.1 Clinton Generating Facility Clinton Nuclear Generating Station (Clinton) is a nuclear facility in southern Illinois operated by Exelon. It was selected as a potential site for conversion to a synchronous condenser due to several indications that it may be required to discontinue operation soon due to changes in economics [9]. According to Walton, a shutdown of the Clinton nuclear facility would cause a loss of almost 1,900 direct and indirect jobs and would make it harder for Illinois to achieve carbon reductions. Clinton is in an area of high wind penetration, where a synchronous condenser is estimated to be very useful due to the lack of inertia from wind generating facilities.

3.2.2 Duane Arnold Generating Facility Duane Arnold Energy Center (Duane Arnold) is a 660 MW nuclear facility located near Cedar Rapids, Iowa and operated by NextEra. Duane Arnold primarily supplies its power to Alliant Energy, and reports indicate Alliant may not renew their contract with Duane Arnold due to their ability to obtain power less expensively through other forms of energy, such as renewables [10]. It is expected that Duane Arnold may close at the end of 2025, nine years before its nuclear generation license expires. Shutdown of the facility is estimated to cause the loss of approximately 600 jobs [11].

3.2.3 Palisades Generating Facility Palisades Nuclear Power Plant (Palisades) is in Michigan and operated by Entergy [12]. The plant has been operating since 1971 and is expected to shut down when its current power purchase agreement expires in 2022. It was selected for study due to the announced shutdown and because it is located close to the Chicago metropolitan area where a synchronous condenser would have a significant amount of impact on the frequency response for the load there.

3.2.4 Prairie Island Generating Facility Prairie Island Nuclear Power Plant (Prairie Island) is in Welch, Minnesota and is operated by Xcel Energy. The facility has been faced with economic challenges which make it hard for the nuclear plant to continue being profitable for Xcel Energy and ratepayers [13]. Prairie island is in an area of high wind penetration and is also very close to the Minneapolis-St. Paul metropolitan area.

3.2.5 Kewaunee Generating Facility Kewaunee Power Station (Kewaunee) was a nuclear generation facility located in Wisconsin that was operated by Dominion Energy [14]. The plant closed in 2013 and much of the existing equipment has been moved.

Table 3 shows a summary of the facilities selected for further analysis along with the estimated size of a synchronous condenser which could be installed at that location based on the overall MVA size of the facility as modeled in the MISO planning models studied.

TABLE 3: MISO NUCLEAR GENERATION FACILITIES SELECTED FOR EVALUATION Synchronous Condenser Nuclear Generation Facility Real Power Output Area Size Estimate (Mvar) Clinton 1188 MW AMIL 1248 Mvar Duane Arnold 660 MW ALTW 721 Mvar Palisades 823 MW METC 894 Mvar Prairie Island Units 1 & 2 593 MW & 593 MW XEL 653 Mvar & 663 Mvar Kewaunee N/A WPS 566 Mvar

The map below shows an overview of the selected locations compared with wind generation facilities [15]. It is expected that synchronous condensers will have a greater impact in areas with high penetration of wind and solar generation projects due to the lack of inertia from those facilities. Since MISO’s primary renewable generation is from wind rather than solar, only the wind penetration is shown in Figure 2.

FIGURE 2: WIND GENERATION WITH SELECTED SITES OVERLAID

Study Procedure To evaluate the frequency response of the system, the following steps were performed.

3.3.1 Create Modified Models Modified versions of the two provided MISO models, summer shoulder and summer peak, were created to represent the retirement of the Clinton, Duane Arnold, Palisades, and Prairie Island nuclear facilities. First, the model of each of these facilities was removed from the base case models provided. The retirements were modeled independently for each of four sites, resulting in four summer shoulder retirement models and four summer peak retirement models. In each model, to modify the load flow parameters, the modeled generation in the same area of the facility was scaled as needed to compensate for the loss of the nuclear real power generation. While in reality the generation would be scaled based on economic and other factors, this approach was selected as a reasonable assumption for determining the frequency response of the facility under retirement conditions.

Modified versions of the models were also created to represent the addition of a synchronous condenser at Clinton, Duane Arnold, Palisades, and Prairie Island to result in another eight study models. The same load flow parameters modified in the retirement case were made to compensate for the loss of the nuclear real power generation. Additionally, the dynamic models were modified to represent a synchronous condenser by decoupling (deactivating) the governor model and adjusting the inertia constant in the generator model to a value near 1 MW-s/MVA.

The model of Kewaunee was created by adding a synchronous condenser model to the 138 kV bus where the Kewaunee facility is located. If a synchronous condenser was installed at Kewaunee, it would likely be installed at a medium voltage level like the other facilities, but modeling at 138 kV was sufficient for purpose of frequency stability analysis. Since the generator parameters of the retired Kewaunee facility are unknown, a sample model similar to the other studied facilities were modified. The model parameters were compared against other synchronous condensers in the MISO model to verify they are similar.

The complete list of models studied is below:

1. Summer shoulder base case (as provided by MISO) 2. Summer shoulder with Clinton retired 3. Summer shoulder with Clinton modeled as a synchronous condenser 4. Summer shoulder with Duane Arnold retired 5. Summer shoulder with Duane Arnold modeled as a synchronous condenser 6. Summer shoulder with Palisades retired 7. Summer shoulder with Palisades modeled as a synchronous condenser 8. Summer shoulder with Prairie Island retired 9. Summer shoulder with Prairie Island modeled as a synchronous condenser 10. Summer shoulder with Kewaunee modeled as a synchronous condenser 11. Summer peak base case (as provided by MISO) 12. Summer peak with Clinton retired 13. Summer peak with Clinton modeled as a synchronous condenser 14. Summer peak with Duane Arnold retired 15. Summer peak with Duane Arnold modeled as a synchronous condenser 16. Summer peak with Palisades retired 17. Summer peak with Palisades modeled as a synchronous condenser 18. Summer peak with Prairie Island retired 19. Summer peak with Prairie Island modeled as a synchronous condenser 20. Summer peak with Kewaunee modeled as a synchronous condenser

3.3.2 Simulate Frequency Disturbances In each case, frequency disturbances were modeled by simulating a large change in real power of the transmission system. For each site studied, a low frequency disturbance was simulated by modeling a large trip in generation in the studied area. The amount of generation to trip was chosen by selecting a large generating facility nearby the plant studied to cause the greatest impact on frequency in the area. A high frequency disturbance was simulated by modeling a large trip in load in the studied area. Table 4 shows the amount of modeled tripped generation and load for each frequency disturbance.

TABLE 4: FREQUENCY DISTURBANCE SIMULATIONS Tripped Generation for Tripped Load for Nuclear Generation Facility Frequency Disturbance Frequency Disturbance All loads in AMIL > 50 MW Clinton 338 MW – total ~1000 MW All loads in ALTW > 20 Duane Arnold 211.9 MW MW for a total of ~500 MW All loads in METC > 25 Palisades 855 MW MW for a total of ~600 MW All loads in XEL > 60 MW Prairie Island Units 1 & 2 675 MW for a total of ~750 MW All loads in WPS > 10 MW Kewaunee 545.8 MW for a total of ~550 MW

3.3.3 Measured Quantities The frequency at the facility high voltage bus was monitored in case to determine the impact on frequency response at that location. The frequency response was determined by measuring these three quantities:

• Frequency Nadir: the minimum (in the case of an underfrequency disturbance) or maximum (in the case of an overfrequency disturbance) frequency obtained after the disturbance was applied, expressed in per-unit variation of 60 Hz. • Time to Nadir: the time after the disturbance was applied at which the frequency nadir was reached, expressed in seconds. • Settling Frequency: the average frequency after the disturbance, measured over the time range of 5 seconds to 10 seconds, showing the short-term steady-state frequency after the disturbance, expressed in per-unit variation of 60 Hz.

All three measurement quantities are displayed in the sample frequency plot in Figure 3.

0.0001 Time to Nadir (seconds) 0.0000 0 1 2 3 4 5 6 7 8 910 -0.0001

-0.0002 Settling Frequency (Hz) Average of recorded frequencies betweeon 5 s and 10 s -0.0003

-0.0004

-0.0005

Frequency Frequency Variation (pu of 60 Hz) -0.0006 Frequency Nadir (pu of 60 Hz) -0.0007 Time (seconds)

FIGURE 3: FREQUENCY MEASUREMENT QUANTITIES An improved frequency response is assumed to be one with a frequency nadir with a smaller absolute value in the synchronous condenser case than in the retirement case or an increased time to reach the frequency nadir. A settling frequency closer to nominal were considered improvements but are not considered a primary effect of an improved frequency response due to a synchronous condenser installation.

3.3.4 Analysis Limitations The analysis shown here was limited to research of the frequency response of only the selected sites. Limitations on the analysis performed include, but are not limited to, the following:

• No modifications were made to the frequency response/controls of wind and solar plants. The response of these facilities is limited to that which was already included in the MTEP summer shoulder and summer peak cases provided by MISO. • Transmission additions or upgrades which have occurred or have been recommended since the MTEP 2016 report study cases were completed are not considered. • Increased wind penetration beyond the amount in the MTEP models provided was not studied. • Study did not look at frequency response of the entire system or local balancing authorities. • Study looked at frequency response in base case models due to a large increase or decrease in frequency due to tripped generation or load. Analysis did not cover the frequency response during a disturbance under contingency conditions. • Voltage stability or other study parameters were not analyzed. • This investigation is not a substitute for thorough planning studies.

4 Results The following pages show the results in each case for a generator trip and a load trip, each in both the summer shoulder and summer peak models. Each modeled disturbance was applied at a simulation time of 1 second and the results are shown out to a simulation time of 10 seconds. An improved frequency response where there is a frequency nadir with a smaller absolute value or an increased time to reach the frequency nadir in the synchronous condenser case as compared to the retirement case are highlighted in yellow.

Clinton Results The analysis results showed that with a synchronous condenser modeled at the site of the Clinton nuclear facility, the frequency response in both the summer shoulder and summer peak cases was improved due to an increase in time to nadir in the summer peak case for both a trip in load and generation. The absolute magnitude of the frequency nadir was negligibly affect in all cases. The results summary for each case is shown in the tables below and the detailed graphs are shown on the following pages.

TABLE 5: CLINTON RESULTS SUMMARY – SUMMER SHOULDER CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00021 -0.00021 -0.00021 0.00034 0.00027 0.00030 Nadir (pu) Time to 0.40002 0.38752 0.40002 0.38752 0.33751 0.46252 Nadir (s) Settling Frequency -0.00006 -0.00006 -0.00006 0.00009 0.00008 0.00008 (pu)

TABLE 6: CLINTON RESULTS SUMMARY – SUMMER PEAK CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00018 -0.00018 -0.00018 0.00030 0.00024 0.00029 Nadir (pu) Time to 0.43752 0.42502 0.43752 0.40002 0.37502 0.46252 Nadir (s) Settling Frequency -0.00005 -0.00006 -0.00006 0.00007 0.00007 0.00007 (pu)

FIGURE 4: CLINTON - GENERATOR TRIP IN SUMMER SHOULDER CASE

FIGURE 5: CLINTON - GENERATOR TRIP IN SUMMER PEAK CASE

FIGURE 6: CLINTON - LOAD TRIP IN SUMMER SHOULDER CASE

FIGURE 7: CLINTON - LOAD TRIP IN SUMMER PEAK CASE

Duane Arnold Results The analysis results showed that with a synchronous condenser modeled at the site of the Duane Arnold nuclear facility, the frequency response in the summer shoulder case was negligibly impacted. The frequency response in the summer peak case was improved. The results summary for each case is shown in the tables below and the detailed graphs are shown on the following pages.

TABLE 7: DUANE ARNOLD RESULTS SUMMARY – SUMMER SHOULDER CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00010 -0.00010 -0.00011 0.00026 0.00018 0.00028 Nadir (pu) Time to 0.20001 0.18751 0.17501 0.23751 0.17501 0.11251 Nadir (s) Settling Frequency -0.00002 -0.00002 -0.00002 0.00004 0.00004 0.00004 (pu)

TABLE 8: DUANE ARNOLD RESULTS SUMMARY – SUMMER PEAK CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00008 -0.00010 -0.00009 0.00048 0.00057 0.00049 Nadir (pu) Time to 0.03751 0.03751 0.03751 0.02501 0.02501 0.02501 Nadir (s) Settling Frequency -0.00001 -0.00001 -0.00001 0.00004 0.00004 0.00004 (pu)

FIGURE 8: DUANE ARNOLD - GENERATOR TRIP IN SUMMER SHOULDER CASE

FIGURE 9: DUANE ARNOLD - GENERATOR TRIP IN SUMMER PEAK CASE

FIGURE 10: DUANE ARNOLD - LOAD TRIP IN SUMMER SHOULDER CASE

FIGURE 11: DUANE ARNOLD - LOAD TRIP IN SUMMER PEAK CASE

Palisades Results The analysis results showed that with a synchronous condenser modeled at the site of the Palisades nuclear facility, the frequency response in both the summer shoulder and summer peak cases was improved with both a decrease in the magnitude of the frequency nadir and an increase in the time to reach the frequency nadir in most studied scenarios. The results summary for each case is shown in the tables below and the detailed graphs are shown on the following pages.

TABLE 9: PALISADES RESULTS SUMMARY – SUMMER SHOULDER CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00042 -0.00039 -0.00037 0.00018 0.00018 0.00018 Nadir (pu) Time to 0.30001 0.28751 0.36251 0.37502 0.35001 0.40002 Nadir (s) Settling Frequency -0.00008 -0.00008 -0.00008 0.00005 0.00005 0.00005 (pu)

TABLE 10: PALISADES RESULTS SUMMARY – SUMMER PEAK CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00038 -0.00036 -0.00035 0.00019 0.00018 0.00019 Nadir (pu) Time to 0.32501 0.30001 0.36251 0.55002 0.57502 0.42502 Nadir (s) Settling Frequency -0.00007 -0.00007 -0.00007 0.00004 0.00004 0.00004 (pu)

FIGURE 12: PALISADES - GENERATOR TRIP IN SUMMER SHOULDER CASE

FIGURE 13: PALISADES - GENERATOR TRIP IN SUMMER PEAK CASE

FIGURE 14: PALISADES - LOAD TRIP IN SUMMER SHOULDER CASE

FIGURE 15: PALISADES - LOAD TRIP IN SUMMER PEAK CASE

Prairie Island Results The analysis results showed that with a synchronous condenser modeled at the site of the Prairie Island nuclear facility, the frequency response in both the summer shoulder and summer peak cases was improved. At Prairie Island, the synchronous condenser case showed more of an improvement as compared to the base case rather than in comparison with the retirement case. This is an unexpected finding and may be worth additional detailed analysis at this site. This result seems to indicate a potential for improvement in frequency response at this location, depending on the parameters of the synchronous condenser or other device which would be installed at this location. The results summary for each case is shown in the tables below and the detailed graphs are shown on the following pages.

TABLE 11: PRAIRIE ISLAND RESULTS SUMMARY – SUMMER SHOULDER CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00081 -0.00072 -0.00074 0.00078 0.00064 0.00072 Nadir (pu) Time to 0.27501 0.23751 0.20001 0.28751 0.30001 0.13751 Nadir (s) Settling Frequency -0.00007 -0.00007 -0.00007 0.00008 0.00008 0.00008 (pu)

TABLE 12: PRAIRIE ISLAND RESULTS SUMMARY – SUMMER PEAK CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00060 -0.00056 -0.00055 0.00094 0.00088 0.00094 Nadir (pu) Time to 0.31251 0.30001 0.27501 0.35001 0.40002 0.43752 Nadir (s) Settling Frequency -0.00006 -0.00006 -0.00006 0.00009 0.00009 0.00009 (pu)

FIGURE 16: PRAIRIE ISLAND - GENERATOR TRIP IN SUMMER SHOULDER CASE

FIGURE 17: PRAIRIE ISLAND - GENERATOR TRIP IN SUMMER PEAK CASE

FIGURE 18: PRAIRIE ISLAND - LOAD TRIP IN SUMMER SHOULDER CASE

FIGURE 19: PRAIRIE ISLAND - LOAD TRIP IN SUMMER PEAK CASE

Kewaunee Results The analysis results showed that with a synchronous condenser modeled at the site of the previously operational Kewaunee nuclear facility, the frequency response in both the summer shoulder and summer peak cases was improved for a generator trip and negligibly impacted for a load trip. The frequency nadir magnitude increased in both cases, but the time to reach the nadir either decreased or stayed the same. The results summary for each case is shown in the tables below and the detailed graphs are shown on the following pages.

TABLE 13: KEWAUNEE RESULTS SUMMARY – SUMMER SHOULDER CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00066 N/A -0.00068 0.00097 N/A 0.00104 Nadir (pu) Time to 0.23751 N/A 0.26251 0.32501 N/A 0.32501 Nadir (s) Settling Frequency -0.00005 N/A -0.00005 0.00007 N/A 0.00007 (pu)

TABLE 14: KEWAUNEE RESULTS SUMMARY – SUMMER PEAK CASE

Generator Trip Load Trip Quantity Synchronous Synchronous Measured Base Case Retirement Base Case Retirement Condenser Condenser Frequency -0.00058 N/A -0.00060 0.00126 N/A 0.00132 Nadir (pu) Time to 0.27501 N/A 0.30001 0.36251 N/A 0.36251 Nadir (s) Settling Frequency -0.00005 N/A -0.00005 0.00008 N/A 0.00008 (pu)

FIGURE 20: KEWAUNEE - GENERATOR TRIP IN SUMMER SHOULDER CASE

FIGURE 21: KEWAUNEE - GENERATOR TRIP IN SUMMER PEAK CASE

FIGURE 22: KEWAUNEE - LOAD TRIP IN SUMMER SHOULDER CASE

FIGURE 23: KEWAUNEE - LOAD TRIP IN SUMMER PEAK CASE

5 Conclusions The results of the thesis reviewed only a handful of potential sites where sites of existing or recently retired nuclear power plants in the MISO operating area could be converted to synchronous condensers. The thesis looked at the existing MISO nuclear facilities of Clinton in Illinois, Duane Arnold in Iowa, Palisades in Michigan, and Prairie Island in Minnesota, as well as the already decommissioned nuclear facility of Kewaunee in Wisconsin. The work completed is seen as a stepping-stone to determine the potential of converting existing conventional power plants to synchronous condensers This section goes through a summary of the results gathered with the analysis of the five facilities studied, the conclusions that were able to be gathered out of those results, and recommendations for future work to be completed in this area.

Summary of Findings A significant improvement in local frequency response was seen at Palisades. This facility is located in an area close to a large amount of load and where there is a large amount of renewable generation projects.

A potential improvement in local frequency response was identified at Prairie Island due to a model response of a decrease in the absolute magnitude of the frequency nadir with a synchronous condenser modeled as compared to in the base case with the Prairie Island nuclear facility in service. Similar to Palisades, Prairie Island is in an area close to a large amount of load and where there is a high amount of wind penetration.

Improvements in local frequency response were also identified at Clinton, Duane Arnold, and Kewaunee, though the improvements were not significant.

All cases had an improvement (i.e. increase) in the time to frequency nadir in either generation trip case or load trip case or both in the case with the synchronous condenser online vs. the simulated retirement scenario.

Conclusions Drawn By Results Converting retired conventional power plants or planning to convert conventional units with pending retirements to synchronous condensers has been determined to be most beneficial in areas that are close to a large amount of load or where there is a large amount of renewable generation projects. At these locations, the rotational inertia provided by the synchronous condensers can provide frequency stability for the grid in a way that renewable energy projects are not able to do with today’s technology.

Future Work The analysis completed as part of this research was focused on local stability in the areas of the five sites studied. Full grid stability is determined by looking at system-wide frequency response, as required by NERC Standard BAL-003-1 [16] which identifies a frequency response obligation (FRO) for an entire balancing authority. Future work is recommending completing a similar analysis but looking at the system-wide frequency stability impacts on converting conventional power plants into synchronous condensers.

The thesis focuses on the impact of converting nuclear facilities only to synchronous condensers, however a similar impact on system stability is expected through the conversion of conventional power plants using other fuels, such as coal, natural gas, or steam. It is recommended to look at the potential of converting conventional power plants of other fuel types to synchronous condensers, as other power plants are also entering retirement at a significant rate as the cost efficiency of renewable energy resources continues to become more competitive.

More detailed analysis is recommended to be completed on specific sites identified as beneficial or otherwise known to be a strong competitor for a synchronous condenser beyond the five locations studied. Furthermore, the study here focuses on those facilities in the MISO region, which generally covers the Midwest and Southeast portions of the United States. Synchronous condenser conversions of nuclear generating facilities and other types of conventional power plants could prove to be beneficial in many other locations.

MISO generation consists of 8% renewables with an increase to up to 30% by 2032 [17]. The analysis completed here utilizes the 2021 MISO system models forecasted from the 2016 MTEP report. The 2021 model includes only the amount of renewable penetration forecasted by MISO to the best of their ability as of the time that study was completed. It would be beneficial to do a similar analysis on a model which has an increased amount of renewable generation. This could be completed on a MISO model with a maximum renewable forecast incorporated, which may be available through MISO as part of another planning study process. It may also be feasible to complete in another area with high renewable penetration like ERCOT in Texas.

References

[1] Nuclear Energy Institute, [Online]. Available: https://www.nei.org/. [Accessed 2018].

[2] NERC, "Standard PRC-024-2 - Generator Performance During Frequency and Voltage Excursions," 2015.

[3] P. E. Marken, A. C. Depoian, J. Skliutas and M. Verrier, "Modern Synchronous Condenser Performance Considerations," San Diego, 2011.

[4] P. Fairley, "Can Synthetic Inertia from Wind Power Stabilize Grids?," IEEE Spectrum, 7 November 2016.

[5] P. Fairley, "Zombie Coal Plants Reanimated to Stabilize the Grid," IEEE Spectrum, 24 July 2015.

[6] J. Eto, J. Undrill, C. Roberts, P. Mackin and J. Ellis, "Frequency Control Requirements for Reliable Interconnection Frequency Response," FERC, February 2018. [Online]. Available: https://www.ferc.gov/industries/electric/indus-act/reliability/frequency-control- requirements/report.pdf. [Accessed 2018].

[7] Midcontinent Independent System Operator, "MISO Corporate Fact Sheet," [Online]. Available: https://www.misoenergy.org/about/media-center/corporate-fact-sheet/. [Accessed 2018].

[8] MISO, "MTEP16: MISO Transmission Expansion Plan".

[9] R. Walton, "MISO: Exelon's Clinton nuclear facility to operate until 2017, but future is hazy," 21 April 2016. [Online]. Available: https://www.utilitydive.com/news/miso-exelons- clinton-nuclear-facility-to-operate-until-2017-but-future-i/417839/.

[10] F. Saunders, "How safe is Iowa's only nuclear plant?," 12 April 2017. [Online]. Available: http://www.thegazette.com/subject/news/business/iowa-nuclear-plant-may-close-in-2025- 20180129.

[11] "Senator: Nuclear power plant closing would be 'nightmare'," 1 February 2018. [Online]. Available: http://wcfcourier.com/news/local/govt-and-politics/senator-nuclear-power-plant- closing-would-be-nightmare/article_4a9991c3-a0f9-5b8f-b6e4-3032b59262c3.html.

[12] R. Adams, "Palisades Nuclear Plant Will Continue Operating Through 2022; Neighbors Ecstatic," 30 September 2017. [Online]. Available: https://www.forbes.com/sites/rodadams/2017/09/30/palisades-nuclear-plant-will-continue- operating-through-2022-neighbors-ecstatic/#3ec410c16ce7.

[13] M. Hughlett, "Xcel seeking new way to charge for nuclear energy costs," Star Tribune, 14 March 2018.

[14] R. Ryman, "Kewaunee County ready to move on after nuclear plant closing," 12 July 2017. [Online]. Available: https://www.lohud.com/story/news/investigations/2017/07/12/kewaunee-nuclear-plant- closing/103598506/.

[15] "AWEA U.S. Wind Industry Map," American Wind Energy Association, 2 December 2017. [Online]. Available: http://gis.awea.org.

[16] NERC, "Standard BAL-003-01 Frequency Response and Frequency Bias Setting".

[17] J. Bakke, D. Duebner, D. Manjure and L. Rauch, "Evolution of the Grid in MISO Region," in MIPSYCON , St. Paul, MN, 2017.