Evaluation of Long Term Power Supply Alternatives February 2019
GDS Associates
Founded in 1986, we serve a client base of municipal and cooperative utility systems with a variety of energy, gas, water and wastewater consulting services, as well as information technology, market research, and statistical services. Headquartered in Marietta, Georgia with 170 employees located throughout the U.S.
Project Team:
Seth Brown, P.E. – Vice President, Transmission ServicesNeil Copeland, P.E. – Managing Director, Power Supply Services John Chiles – Principal, Transmission ServicesDr. Bill Jacobs, P.E. – Executive Engineer – Generation Services
2
Background
Evaluation of Power Supply Alternatives
Study Objective: Evaluate long-term power supply alternatives including Nuclear Development – Bellefonte Project Power Purchase AgreementCost of Energy-only modeledEvaluate MLGW as both stand-alone and integrated into MISO2022 Study Year Include 15% renewable (wind) portfolioCompare to current TVA wholesale power agreement – NOTE THAT
STUDY DID NOT INCLUDE VALUE/COST OF CAPACITY OR COSTS ASSOCIATED WITH ANY NEW DEBT SERVICE
3
Objective
Methodology & Principal Assumptions
Energy Methodology & Principal Assumptions
Utilized a large footprint (excludes Florida, New England, NE Canada, and Saskatchewan) containing load, generation, and nodal modeling (substation level analysis)Full nodal analysis using PROMOD IV production cost software and
the latest MISO database for the Calendar Year of 2022Produces a fully integrated, security constrained economic dispatch
adhering to generation and transmission limitations simultaneouslyCaptures unit generation metrics, transmission congestion, and load
costs. Does not include capacity costs/valueTVA Business-As-Usual Case represents continuation of current
wholesale power agreement that includes capacity costs. PROMOD results for TVA fleet include production costs only (fuel + operations & maintenance)
Utilized a large footprint (excludes Florida, New England, NE Canada, and Saskatchewan) containing load, generation, and nodal modeling (substation level analysis)Full nodal analysis using PROMOD IV production cost software and
the latest MISO database for the Calendar Year of 2022Produces a fully integrated, security constrained economic dispatch
adhering to generation and transmission limitations simultaneouslyCaptures unit generation metrics, transmission congestion, and load
costs. Does not include capacity costs/valueTVA Business-As-Usual Case represents continuation of current
wholesale power agreement that includes capacity costs. PROMOD results for TVA fleet include production costs only (fuel + operations & maintenance)
5
Methodology
Energy Methodology & Principal Assumptions
In the “MLGW BA” scenarios, a production cost construct is used (production costs + purchases – sales) These loads include losses in the production costs calculations In the “MLGW MISO” scenarios, a market based approach using
Locational Marginal Pricing (LMP) is implemented (generator margins offset load costs) System is dispatched to load with losses, but MLGW only pays substation loads Generators participating in the market are paid a loss component of LMP to
compensate for their additional generation to cover actual losses on the system Four additional renewable scenarios across the four primary scenarios
included power purchase agreements of 700 MW of high capacity factor wind from the upper midwestern portions of MISOWind energy provided approximately 15% of MLGW energy
requirements
In the “MLGW BA” scenarios, a production cost construct is used (production costs + purchases – sales) These loads include losses in the production costs calculations In the “MLGW MISO” scenarios, a market based approach using
Locational Marginal Pricing (LMP) is implemented (generator margins offset load costs) System is dispatched to load with losses, but MLGW only pays substation loads Generators participating in the market are paid a loss component of LMP to
compensate for their additional generation to cover actual losses on the system Four additional renewable scenarios across the four primary scenarios
included power purchase agreements of 700 MW of high capacity factor wind from the upper midwestern portions of MISOWind energy provided approximately 15% of MLGW energy
requirements
6
Methodology
Energy Methodology & Principal Assumptions
Scenario A: MLGW as its own BA w/ BellefonteBellefonte is delivered to MLGW via Firm PtP TransmissionMLGW holds Firm PtP to MISO for peak load (loss of Bellefonte) Scenario B: MLGW as its own BA w/ Bellefonte and new self-build resourcesBellefonte is delivered to MLGW via Firm PtP TransmissionMLGW holds hourly non-Firm service to and from MISO for sales and
purchases Scenario C: MLGW in MISO w/ BellefonteBellefonte is delivered to MISO via Firm Point to Point (PtP) TransmissionMLGW holds Firm PtP to MISO for peak load (Pseudo-Tie and loss of
Bellefonte) Scenario D: MLGW in MISO w/o BellefonteMLGW holds Firm PtP to MISO for peak load (Pseudo-Tie)Procures all energy from MISO
Scenario A: MLGW as its own BA w/ BellefonteBellefonte is delivered to MLGW via Firm PtP TransmissionMLGW holds Firm PtP to MISO for peak load (loss of Bellefonte) Scenario B: MLGW as its own BA w/ Bellefonte and new self-build resourcesBellefonte is delivered to MLGW via Firm PtP TransmissionMLGW holds hourly non-Firm service to and from MISO for sales and
purchases Scenario C: MLGW in MISO w/ BellefonteBellefonte is delivered to MISO via Firm Point to Point (PtP) TransmissionMLGW holds Firm PtP to MISO for peak load (Pseudo-Tie and loss of
Bellefonte) Scenario D: MLGW in MISO w/o BellefonteMLGW holds Firm PtP to MISO for peak load (Pseudo-Tie)Procures all energy from MISO
7
Scenarios
Energy Methodology & Principal Assumptions
GDS used NYMEX HH futures from 9/25/2018Represents price of fuel burned by generators in the TVA region, and
the new MLGW generationCoal prices represent the average coal prices from all coal plants in
TVA
GDS used NYMEX HH futures from 9/25/2018Represents price of fuel burned by generators in the TVA region, and
the new MLGW generationCoal prices represent the average coal prices from all coal plants in
TVA
8
Fuel
$3.32 $3.28 $3.20 $2.91 $2.83 $2.76 $2.79 $2.80 $2.93 $2.96 $3.08 $3.22
($0.50)
$0.00
$0.50
$1.00
$1.50
$2.00
$2.50
$3.00
$3.50
$4.00
$/MMBtu
2022 Monthly Fuel PricesBurner Tip Adder TX Gas Z3 Henry Hub Average Coal Price
Market
Generation Supply
The MISO market is well-supplied with generation and carries greater than 27% reserve margins (including non-firm load), allowing opportunities for interchange between control areas In 2022, combined cycle generating capacity is approximately 29,000 MW Low cost generation should be able to generate energy margins to offset load
costs
The MISO market is well-supplied with generation and carries greater than 27% reserve margins (including non-firm load), allowing opportunities for interchange between control areas In 2022, combined cycle generating capacity is approximately 29,000 MW Low cost generation should be able to generate energy margins to offset load
costs
10
MISO Supply Stack
Energy Results
TVA Business as Usual
PROMOD yields a TVA “energy only” cost of production of $18.50/MWh TVA SEC 10-K Filing indicates the full-delivered, all-requirements cost to
serve MLGW is ~$66.00/MWh
12
Costs
Scenario A
This “MLGW BA” scenario incurs the Bellefonte costs with no revenue offset Imports from MISO make up the balance of the native load requirements Import costs are priced at the load-weighted MLGW LMP to be consistent
with the “MLGW MISO” scenarios
This “MLGW BA” scenario incurs the Bellefonte costs with no revenue offset Imports from MISO make up the balance of the native load requirements Import costs are priced at the load-weighted MLGW LMP to be consistent
with the “MLGW MISO” scenarios
13
Costs
$(439)
$(735)
$‐
$(178) $0 $(34)
$(85) $(800)
$(700)
$(600)
$(500)
$(400)
$(300)
$(200)
$(100)
$‐
Therm Gen Non‐Therm Gen Imports Exports Gen PtP Load PtP Net Costs
Millions
Scenario A
Scenario A (Wind)
As compared to Scenario A, imports from MISO are lower due to the purchase of renewable wind energyThe reduction in import costs from the market are offset by the price
of wind and the transmission costs to move the wind through MISONet costs are almost identical to Scenario A without the wind
As compared to Scenario A, imports from MISO are lower due to the purchase of renewable wind energyThe reduction in import costs from the market are offset by the price
of wind and the transmission costs to move the wind through MISONet costs are almost identical to Scenario A without the wind
14
Costs
$(439)
$(735)
$(41)
$(126) $11 $(54)
$(85) $(800)
$(700)
$(600)
$(500)
$(400)
$(300)
$(200)
$(100)
$‐
Therm Gen Non‐Therm Gen Imports Exports Gen PtP Load PtP Net Costs
Millions
Scenario A (Wind)
Scenario B
This “MLGW BA” scenario includes the Bellefonte, 2 new CC’s, and 6 new CT’s (new capacity of 2,606 MW), and provides a 16% planning reserve margin Increased exports provide additional revenue opportunitiesMLGW holds non-firm hourly transmission service since their load is
covered by internal generation
This “MLGW BA” scenario includes the Bellefonte, 2 new CC’s, and 6 new CT’s (new capacity of 2,606 MW), and provides a 16% planning reserve margin Increased exports provide additional revenue opportunitiesMLGW holds non-firm hourly transmission service since their load is
covered by internal generation
15
Costs
$(637) $(642)$‐ $(16)$44 $(34) $‐
$(800)
$(700)
$(600)
$(500)
$(400)
$(300)
$(200)
$(100)
$‐
Therm Gen Non‐Therm Gen Imports Exports Gen PtP Load PtP Net Costs
Millions
Scenario B
Scenario B (Wind)
Using the wind to serve load provides more opportunities for additional generation exports as compared to Scenario B Increased energy margins cannot overcome the cost of procuring and
transporting the wind across the MISO systemNet costs are $23MM higher when compared to Scenario B without the
wind
Using the wind to serve load provides more opportunities for additional generation exports as compared to Scenario B Increased energy margins cannot overcome the cost of procuring and
transporting the wind across the MISO systemNet costs are $23MM higher when compared to Scenario B without the
wind
16
Costs
$(598)$(665)$(41) $(10)
$55 $(72) $‐
$(800)
$(700)
$(600)
$(500)
$(400)
$(300)
$(200)
$(100)
$‐
Therm Gen Non‐Therm Gen Imports Exports Gen PtP Load PtP Net Costs
Millions
Scenario B (Wind)
Scenario C
In this “MLGW MISO” scenario Bellefonte receives revenue from the MISO market, which offsets some of the PPA costs PtP transmission costs are incurred to integrate into the MISO market since
MLGW has no direct electrical transmission connections to MISO
In this “MLGW MISO” scenario Bellefonte receives revenue from the MISO market, which offsets some of the PPA costs PtP transmission costs are incurred to integrate into the MISO market since
MLGW has no direct electrical transmission connections to MISO
17
Costs
$237
$(754)
$(439) $‐ $‐ $(34)$(83)
$(435) $(1,000)
$(800)
$(600)
$(400)
$(200)
$‐
$200
$400
Ther Rev Ther Cost Non‐Ther Rev Non‐Ther Cost Gen PtP Load PtP Load Costs Net Costs
Millions
Scenario C
Scenario C (Wind)
With MLGW and the wind in MISO, MLGW would pay the congestion costs to move the wind across the MISO system as compared to firm PtP in the MLGW BA wind scenarios Hourly load is reduced by the hourly schedules of the wind, which is then
multiplied by the MLGW hourly load-weighted LMP to arrive at the Load Costs Net costs are $14MM higher than Scenario C without the wind
With MLGW and the wind in MISO, MLGW would pay the congestion costs to move the wind across the MISO system as compared to firm PtP in the MLGW BA wind scenarios Hourly load is reduced by the hourly schedules of the wind, which is then
multiplied by the MLGW hourly load-weighted LMP to arrive at the Load Costs Net costs are $14MM higher than Scenario C without the wind
18
Costs
$242
$(768)
$(439) $‐ $(41) $(34)$(83) $(42)
$(372) $(1,000)
$(800)
$(600)
$(400)
$(200)
$‐
$200
$400
Ther Rev Ther Cost Non‐Ther Rev Non‐TherCost
Gen PtP Load PtP Cong Costs Load Costs Net Costs
Millions
Scenario C (Wind)
Scenario D
This “MLGW MISO” scenario purchases all energy from the MISO market, and shows the lowest energy costs of all scenariosWith no hedges, MLGW would be subject to market scarcity energy
pricing
This “MLGW MISO” scenario purchases all energy from the MISO market, and shows the lowest energy costs of all scenariosWith no hedges, MLGW would be subject to market scarcity energy
pricing
19
Costs
$‐
$(529)
$‐ $‐ $‐ $‐$(83)
$(445)
$(1,000)
$(800)
$(600)
$(400)
$(200)
$‐
Ther Rev Ther Cost Non‐Ther Rev Non‐Ther Cost Gen PtP Load PtP Load Costs Net Costs
Millions
Scenario D
Scenario D (Wind)
The purchase of wind lowers the Load Costs, but is not able to overcome the cost and transportation of the wind across the MISO systemNet costs are $18MM higher than Scenario D without the wind
The purchase of wind lowers the Load Costs, but is not able to overcome the cost and transportation of the wind across the MISO systemNet costs are $18MM higher than Scenario D without the wind
20
Costs
$‐
$(547)
$‐ $‐ $(41) $‐$(83) $(43)
$(381)
$(1,000)
$(800)
$(600)
$(400)
$(200)
$‐
Ther Rev Ther Cost Non‐Ther Rev Non‐TherCost
Gen PtP Load PtP Cong Costs Load Costs Net Costs
Millions
Scenario D (Wind)
Summary of Scenarios
21
Costs
$735.0
$665.4
$768.1
$547.5
$735.3
$641.9
$754.1
$528.6
$946.4
0 100 200 300 400 500 600 700 800 900 1,000
Scenario A
Scenario B
Scenario C
Scenario D
TVA WPA (2017)
$ Mil
Scenario CostsNo Wind Wind
TVA’s “all‐in” cost to serve
MLGW…Scenarios do not include all costs to serve
MLGW.
Summary of Scenarios
Bellefonte costs are well above market energy prices under modeled gas prices. Comparison of MISO scenarios (D minus C) shows a ($200MM) differential owning Bellefonte in MISO vs MISO-only. Bellefonte and TVA provide a capacity benefit. New, efficient thermal generation provides hedges against market prices, and
should provide energy margins to offset load costs, but requires capital Purchasing strictly from the market provides opportunities for low-cost power,
but provides no protection from scarcity energy pricing. Capacity can be procured from the MISO market but prices fluctuate annually
Bellefonte costs are well above market energy prices under modeled gas prices. Comparison of MISO scenarios (D minus C) shows a ($200MM) differential owning Bellefonte in MISO vs MISO-only. Bellefonte and TVA provide a capacity benefit. New, efficient thermal generation provides hedges against market prices, and
should provide energy margins to offset load costs, but requires capital Purchasing strictly from the market provides opportunities for low-cost power,
but provides no protection from scarcity energy pricing. Capacity can be procured from the MISO market but prices fluctuate annually
22
Costs
Generation Supply
23
MISO Supply Stack
62,585 126,650 99,755
MLGW CC 1
MLGW CT 1
Bellefonte …
TVA’s “all‐in” cost to serve MLGW
TVA Energy Only
$0
$10
$20
$30
$40
$50
$60
$70
0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000
Dispa
tch Price ($/M
Wh)
Cumulative Capacity (MW)
MISO Supply Stack (July 2022)
Min LoadMin LoadMax LoadMax Load
AverageLoad
AverageLoad
Bellefonte Project
Issues associated with Bellefonte Project viabilityFramatome’s technical expertise with this reactor designMany original equipment vendors no longer in existence requiring
reverse engineering of componentsLack of a detailed engineering analysis of the existing plant systems
and equipmentUse of Maximum Guaranteed Price (MGP) contracts with penalties
assessed to the contractors for schedule delays may be unrealisticProgressing from fuel load to commercial operation in three months
may be unrealisticAbility to hire and train operators and development of a plant
simulator may be problematic
24
Risks
Possible Next Steps
Obtain data from TVA on the incremental cost of capacity, energy, transmission, and ancillary services required to serve MLGWConduct a “discovery session” with MISO Identify transmission transfer limitations with TVA and MISO Review the need to develop an Integrated Resource Plan: Identify corporate goals for renewables, demand response and energy
efficiency Identify the most cost-effective resource portfolioDevelop long-range financial forecast associated with new resources (cash-
flow impacts, debt service limitations, financial metrics, customer rate impacts)
Conduct a Request for Proposals for new resources
25
Conclusion