Item 14
VALLEY CLEAN ENERGY ALLIANCE
Staff Report – Item 14
____________________________________________________________________________________ TO: Board of Directors
FROM: Gordon Samuel, Assistant General Manager & Director of Power Services SUBJECT: Consider increasing VCE’s Existing 80% Renewable by 2030 Goal DATE: July 13, 2023
RECOMMENDATIONS
1. Receive presentation and provide feedback on VCE’s power portfolio content goals.
2. Increase the current 80% renewable goal by 2030 to 100% renewable by 2030 and substitute the 25% renewable local component goal with a goal of 25% of future storage amounts to be from local installations.
3. When conducting solicitations state a preference for locally sited resources.
BACKGROUND
Prior to VCE’s launch in June 2018, the Board adopted a goal for VCE’s power content to target 80% renewables by 2030. The goal also set a requirement that 25% of this amount should be from local resources. At the time this was a very ambitious goal, and some may still consider this to be a stretch or at least a sufficient target. Others may believe this goal does not go far enough. Since this goal was adopted, VCE has entered into several long-term power purchase agreements (PPAs) and has been working towards fulfilling these goals.
Guiding Documents – Carbon Neutral Study and 2022 Integrated Resource Plan In the second half of 2021 and early 2022, staff completed a 100% carbon neutral by 2030 study (CNx2030) which considered not only being carbon neutral but also 100% renewable (100% Carbon Free Portfolio Study (Final). In 2022, VCE submitted its integrated resource plan (IRP) to the California Public Utilities Commission (CPUC) (Integrated Resource Plan (IRP) filed 11/1/2022). This IRP also studied various portfolios from 2023-2035 with the primary focus to be at or below a specific emissions target in an effort to reduce greenhouse gases (GHG) at the lowest cost. As a result of VCE’s procurement and study efforts, a reasonable roadmap is beginning to emerge which presents an opportunity to revisit the current power content goal adopted in 2018.
During the November 17, 2022 Community Advisory Committee (CAC) meeting, the Committee voted unanimously to recommend that the Board modify the existing goal. The CAC recommended that the Board approve a new goal which is 100% renewable by 2030 with 25% of the content sourced from local resources. A key point expressed by several CAC members who spoke in support of increasing the goal believed that it provides a reasonable target for VCE to aspire towards. Unfortunately, in November 2022, Staff did not have the supporting analysis of the advantages/disadvantages associated 2 with the CAC proposed modification of the current portfolio goal – specifically the 25% local component. Staff reported this CAC recommendation to the Board in late 2022 and received direction from the Board to return in mid-2023 with an analysis of setting a more aggressive 2030 renewable target.
Based on that direction, staff engaged the portfolio modeling services of First Principles Advisory (FPA), the firm that performed the portfolio modeling for VCE’s 2022 IRP. FPA’s familiarity with the VCE portfolio was a logical reason to have them conduct this additional modeling. At the February 2023 Board meeting, the Board approved a contract with FPA for the additional modeling with a commitment from staff to bring the results and recommendation to the CAC in June and the Board in July 2023.
VCE Current Renewable Portfolio
VCE’s has signed seven renewable PPA consisting of photovoltaic (PV), hybrid (PV + storage) and geothermal. These PPAs account for approximately 680 annual GWHs or approximately 70% of VCE’s retail load (2030).
Table 1 – VCE’s Executed Long-Term Renewable PPAs
Long Term PPAs Actual or Expected COD Capacity*
Resurgence Solar I 7/7/2023 (active) 90 MW PV, 75 MW BESS (250,000 MWhs)
Aquamarine Solar 9/22/2021 (active) 50 MW PV (130,000 MWhs) Putah Creek Energy Farm 10/15/2022 (active) 3 MW PV, 3 MW BESS (7,600 MWhs) Gibson Solar 6/1/2025 13 MW PV, 13 MW BESS (50,000 MWhs) Willy 9 Chap 2** 12/31/2023 72 MW PV, 36 MW BESS (210,000 MWhs)
Ormat Geothermal Varies by project, but as early as
2025 4.63 MW (35,380 MWhs)
Fish Lake Geothermal June 2024 0.42 MW (3,460 MWhs)
* All Battery Energy Storage Systems (BESS) are 4-hour duration, except the Gibson Solar project is a 5-hour battery system. Approx annual MWhs shown.
** Formerly Willow Springs Solar 3. Name changed at the request of the CAISO.
ANALYSIS
Staff, working with FPA, identified seven scenarios to model. Four assuming a future natural gas price curve that would be considering a P50 curve and three at a higher price P95 curve1. For the purposes of this report, staff elected to focus on the results associated with the P50 analysis. Note: if natural gas prices are more in-line with the P95 assumption this does have a material cost impact on the portfolio depending on the type of renewable resources selected.
1 P50 and P95 represent the confidence level of a cost not being exceeded. A P50 value has a 50% probability that it will be exceeded, whereas, a P95 only has a 5% probability of being exceeded. 3 Table 2 – Four Scenarios Modeled Based on P50 NG Curve Scenario NG Price 2030 RPS
Target
Local RPS
Target
1.a (current goal) P50 80% 25%
1.b P50 100% 25%
1.c P50 100% -
1.d P50 80% -
Scenario 1.a can be considered the base case or business as usual (BAU). This scenario is the current VCE goal. Scenario 1.b increases the renewable percentage to 100% by 2030 as well as maintains the current 25% local renewable component. Scenario 1.c increases the renewable percentage to 100% but only considers the two existing local PPAs that VCE has executed (Putah Creek and Gibson2).
Finally, Scenario 1.d maintains the current 80% renewable by 2030 and only considers the two existing local PPAs that VCE has executed (Putah Creek and Gibson). The purpose of identifying these four scenarios was to establish a range of potential cost outcomes.
The portfolios from each of the scenarios are slightly different but the primary choice of eligible renewable technologies does not vary (note: the modeling does allow for the selection of other technologies such as biomass, off-shore wind, etc but only selects resources that are the best fit for the portfolio). Table 3 below identifies the incremental capacity additions (additions above what VCE has already contracted) for each scenario.
Table 3 – Cumulative Incremental Capacity Additions
As shown the local hybrid renewable resources are assumed to be PV + storage as that is the most realistic resource available in Yolo County as the county does not have significant geothermal or wind resources and the local biomass resources have proven to be quite costly. Unfortunately, PV land 2 These two local projects represent approximately 8% of VCE’s renewable portfolio. 4 usage is significant which can have impacts on prime agricultural land which presents policy trade-offs and can be difficult to permit. To achieve the current 25% local goal, the amount of new local capacity would be between 65 MW and 84 MW. Although this amount is technically feasible, based on direct experience staff observes that it will be difficult to permit locally, the costs will be at a premium compared to installation in other regions of the State, and it installs a technology that VCE otherwise would not select as the portfolio would benefit from additional diversification (e.g. wind).
Scenario Costs
Table 4 below identifies the net present value (NPV) cost trade-offs between each scenario. All scenarios are measured off Scenario 1.a (BAU scenario). Scenario 1.b is $23.5 million more than the BAU case, similarly Scenario 1.c is $33.4 million cheaper than the BAU case (or nearly $57 million cheaper than Scenario 1.b). Scenario 1.d is the lowest cost of all cases. An important point to highlight is the incremental cost to go from 80% (1.d) to 100% (1.c) renewable is not unreasonable and staff believes this is something the Board should consider. It is clear from the model runs that portfolio costs are amplified when factoring in additional local resources.
Table 4 – Scenario Cost Comparisons
Goals / Policies of other Load Serving Entities (LSEs) in California – Including CCAs Although each LSE’s situation is different, it is important to understand what other LSE’s have committed to. Numerous LSEs do have “aspirational” goals of achieving 100% renewable and clean power by 2030. Many LSEs intentionally include the term “clean” in their goals as this allows some flexibility to meet some of the content with resources such as large hydro or nuclear (both are defined as GHG-free or clean but neither qualify as renewable per the CPUC definition). For reference, Attachment 1 identifies the policies of many LSEs in California.
Strategic Plan
VCE’s current Strategic Plan contains the following goal: “Manage power supply resources to consistently exceed California’s Renewable Portfolio Standard (RPS) while working toward a resource portfolio that is 100% carbon neutral by 2030.” By definition, California RPS renewable energy is also carbon free. Therefore, technically VCE would be exceeding its strategic plan goal by modifying to a 100% renewable goal by 2030. In addition, aspects of the strategic plan are currently being reviewed so any new goals can be incorporated into the latest version.
Scenario NG Price
2030 RPS
Target
Local RPS
Target
2024-2035
NPV
(2022 $M)
Delta
(2022 $M)
1.a P50 80% 25% 619.6 0
1.b P50 100% 25% 643.1 23.5
1.c P50 100% - 586.1 -33.4
1.d P50 80% - 575.7 -43.9
5 Community Advisory Committee Feedback
Staff presented this topic to the CAC in June. The topic generated a robust discussion of the pros and cons of increasing the goal to 100% as well as the local component modification. The CAC was supportive of Staff’s recommendation and further requested a slight enhancement of the recommendation. The CAC proposed (and staff supports), that when VCE conducts solicitations it states there is a preference for locally sited resources. In addition, the CAC believes the definition of “local” to be expanded to include Yolo County and the adjacent counties.
CONCLUSION
Staff is seeking Board feedback on modifying the existing power portfolio goal. Staff findings based on the scenario analysis:
1) With the current portfolio trajectory staff believes achieving 100% renewable is a reasonable and fiscally sound goal for the Board to consider.
2) Staff also believes the 25% local component should be revisited for two reasons:
a. The primary local renewable resource is solar (PV). From the modeling, additional solar is not a resource VCE needs as VCE needs to diversify to other renewable technologies to achieve a more balanced renewable portfolio.
b. The cost to achieve the 25% local requirement should be considered as this decision is discussed. It is substantially more costly to VCE’s customers to meet this component of the goal by 2030 and beyond. Based on the analysis, savings of up to $33M may accrue over the next 10 years vs. BAU (see Table 4).
3) Stand-alone storage resources, which enable the installation of more intermittent renewable resources in California (e.g. solar, wind), is a resource that could be considered for Yolo County. a. Permitting stand-alone storage is likely to prove to be more stream-lined as the footprint of the underlying disturbed land is much smaller than solar. b. VCE, as well as the grid, need storage.
Attachments
1. California LSE’s renewable goals
2. Study Results from First Principles Advisory (redacted) 6 Attachment 1
Renewable Energy Goals of CCAs and other electric utilities in California (as of May 2023) LSE Name Goal Summary
IOUs
PG&E net-zero energy system by 2040
SCE RPS
SDG&E RPS
Regional/Municipal Districts
IID (Imperial Irrigation
District) RPS
LDWP 80%/90% RPS by 2030, 100% carbon-free by 2035
SMUD zero carbon emission power supply by 2030
CCAs
Apple Valley Choice
Energy RPS
Central Coast
Community Energy 100% clean and renewable energy by 2030 City of Palmdale RPS
City of Pomona RPS
City of Santa Barbara 75% renewable / 100% carbon-free by 2030 (default rate) Clean Energy Alliance 100% renewable by 2035
Clean Power Alliance RPS
CleanPowerSF 100% renewable electricity by 2025, and 100% renewable energy (0% fossil fuels) by 2040
Desert Community
Energy RPS plus 100% carbon-free product as default for customers in Palm Springs East Bay Community
Energy
100% net-zero carbon annually by 2030, 100% clean energy on a net annual basis by 2030, exceed state RPS by 20% per year
King City Community
Power RPS
Lancaster Choice
Energy RPS
Marin Clean Energy 60% minimum renewable, default rate at 98.3% renewable as of 2021 (including large hydro)
Orange County Power
Authority RPS
Peninsula Clean Energy 100% renewable by 2025, including hourly matching (i.e., time-coincident basis) Pico Rivera Innovative
Municipal Energy 50% renewables (default rate)
Pioneer Community
Energy RPS
7 Rancho Mirage Energy
Authority RPS
Redwood Coast Energy
Authority
100% clean and renewable by 2025, plus 100% local renewable energy by 2030 (local resources located within the Humboldt Local Capacity Area) San Diego Community
Power
75% in 2027, 85% in 2030, and 100% in 2035 renewable; 15% new storage in Member Agencies' territories by 2035; 600MW of new utility scale projects within San Diego and Imperial Counties by 2035
San Jacinto Power RPS
San Jose Clean Energy 100% carbon neutral and renewable (annual basis) by 2030, with 0% fossil fuel by 2050
Silicon Valley Clean
Energy Authority
50% - 52% renewable currently, 60% - 62% renewable in 2030; 100% of energy needs with carbon-free electricity on annual basis, longer-term goal of carbon-free on 24x7 basis
Sonoma Clean Power Board policy of 50% renewable by 2020; Internal goal of 100% hourly marginal emissions mitigation by 2026 and 80% Winter Night Reliability by 2030 Valley Clean Energy
Alliance. 80% renewable by 2030
8 Attachment 2 – Study Results from First Principles Advisory (redacted) VCE Modeling Results
Spring 2023 VCE Spring 2023 Modeling Exercise
Contents
• Executive Summary
• Key Recommendations
• Financial Summary
• Local Project Assumptions
• Detailed Breakdown of Model Results for Scenario 1.a
• System Buildout / Annual Generation / Resource Adequacy / ELCCs • Hourly Dispatch
• Market Participation
• Portfolio Costs
• RPS
• Sensitivity Case: Scenario 3 (capped market transactions) • Procurement Implications
• Recommendations
• Technology Breakeven Prices
• Potential Next Steps
• Key Modeling Inputs / Assumptions
• VCE load
• RA forecast / NG forecast / LMP forecast
• Technology cost curves / ELCC curves
• Model Related Info (description of process, key assumptions, Gridpath comparison to MATCH) Executive Summary
• VCE’s current baseline scenario (i.e., 80% RPS energy by 2030 with a 25% local carveout) increases the NPV of the portfolio by $20-64M across the 2024-2035 planning horizon • Estimated cost premium is dependent on multiple variables, including (but not limited to) the price of NG and VCE’s risk management policy regarding market transactions
• Increasing the RPS target to 100% and keeping the 25% local carveout will result in an additional NPV expense of $14-24M relative to the baseline scenario
• Assuming solar-storage hybrid projects is the only available technology that can be built in Yolo County, VCE’s portfolio heavily is concentrated in solar and 4-hr storage and has limited ability to diversify its supply with other technologies
• Onshore wind, geothermal, and storage greater than 4 hours of duration constitute a greater share of contracted supply for the scenarios where the local constraint is not active • Combined, the recently passed IRA legislation and the current strained market conditions for li-ion based storage projects have significantly increased the opportunity cost associated with VCE’s local generation policy
• Li-ion costs have increased 15-25% whereas cost of alternatives have decreased 20-30% • Despite current high RA-only prices, VCE will be better off if they wait for market conditions for li-ion storage to settle and not lock in high price long-term (i.e. 10yr+) contracts • Note interconnection and project constructions delays increase the complexity of the near-term procurement strategy Key Recommendations
• 100% (annual) RPS target by 2030 is economically feasible • Policy would not add significant costs to the portfolio • Allowing stand-alone storage projects located in Yolo County to qualify can help reduce the associated cost premium
• Path dependency matters -> need to know the end destination and what’s allowed; update as you go
• A clearly defined RPS / GHG policy and risk management policy is necessary to ensure portfolio is heading in the desired direction
• Understanding of current PPA market is critical
• Run an all-source solicitation annually to obtain timely information on technology costs • Execute agreements when favorable prices of desired technologies present themselves • Curtailment risk is only partially accounted for in current modeling approach • A site-by-site assessment should be taken in future modeling exercises • Benefits of customer programs should be fully explored to contain costs and increase flexibility • Updated TEA load forecast assumes VCE grows its annual load by 35% while increasing its peak demand by only 18% Financial Summary
• In the P50 NG scenario, increasing VCE’s 2030 RPS target and keeping the 25% local requirement results in an additional charge of $23.5M in present value relative to VCE’s current policy (e.g., 80% RPS with a 25% carve out) • If VCE increases their 2030 RPS target to 100% but doesn’t require any additional local projects results in a cost savings of $33.4M • In the P95 NG scenario, the local premium is $21.6M
Index Natural
Gas Price
RPS 2030
Target
Local RPS
Requirement
Local Hybrid
Capacity
(MW)
2024-2035
NPV
(2022 $M)
Delta
(2022 $M)
1.A P50 80% 25% 65 619.6 0
1.B P50 100% 25% 84 643.1 23.5
1.C P50 100% n/a 0 586.1 -33.4
1.D P50 80% n/a 0 575.7 -43.9
2.A P95 80% 25% 65 591.6 0
2.B P95 100% 25% 84 613.2 21.6
2.C P95 100% n/a 0 527.4 -64.3
*Scenario 2.D not run in GridPath because model doesn’t bind on RPS constraint in scenario 2.C Local Project Characteristics
• 1MW-4MWh of storage for every 1 MW
of solar
• Solar capacity factor: 33%
• Battery roundtrip efficiency: 86%
• Local hybrid project is ~40% more
expensive than latest technology costs
curves posted in CPUC’s latest Inputs and
Assumptions (I&A) document
Year 80% RPS
Target
100% RPS
Target
2028 22.5 30
2030 45 60
2032 55 72
2035 65 84
Cumulative Capacity (MW)
Year 2028 2030 2032 2035
Energy Price ($/MWh) 54 50 50 45
Capacity Price ($/kW-m) 10.75 9.50 8.00 7.00
“All-in Price”
($/MWh)
100 90 84 75
Non-Local Hybrid All-in
Price ($/MWh)
70 64 60 54
Detailed Results of Scenario 1.a
Scenario 1.A: System Buildout (nameplate)
• Solar (stand-alone or hybrid) makes up 73% (60%) of
nameplate capacity in 2030 (2035)
• Hybrid storage makes up 67% (54%) of nameplate
capacity in 2030 (2035)
• Hybrid storage is 4 hours in duration
• Dashboard File:
• MW
• Portfolio Summary I: columns B:H
• Portfolio Summary II: columns D:J
• MWh
• Portfolio Summary I: columns L:R
• Portfolio Summary II: columns P:V
Scenario 1.A: Generation (contracted)
• Solar (stand-alone or hybrid) makes up
75% (65%) of contracted generation in
2030 (2035)
• Curtailment risk of VER resources only
partially addressed in this analysis;
additional analysis warranted
• Dashboard File:
• MWh
• Portfolio Summary I: columns V:AB
• Portfolio Summary II: columns AB:AH
Scenario 1.A: ELCC – Resource Adequacy
• Solar (stand-alone or hybrid)
makes up 63% (47%) of firm
capacity in 2030 (2035)
• The more local capacity built into
the portfolio, the less opportunity
for other technologies types to be
added
• Model calls on longer duration
storage in part to benefit from
larger ELCCs to satisfy the RA
constraint
• Dashboard File:
• Worktab: RA Constraint
• Firm ELCC: columns N:T
• Breakdown of ELCCs: columns AI:AL
• Adjust cell AF1
2030 2035
Scenario 1.A: Hourly Dispatch (2024)
Jan Apr
Jul Oct
• The gap between the white
line (busbar_load) and the
black line (total_supply)
represents market activity
• Purchases are when load
is greater than supply
• Sales are when supply is
greater than load
Dashboard File:
Worktab:
dispatch_tech_total
Pivot Table: columns CP:DF
Note: be sure to refresh
table when updating year of
interest on
meet_load_constraint tab
Scenario 1.A: Hourly Dispatch (2030)
Jan Apr
Jul Oct
• VCE is a net buy during the
early morning hours
throughout the year
• VCE is a net seller for the day
and most of the evening in
non-winter months
• Max sales are during
morning and afternoon
peaks
Dashboard File:
Worktab:
dispatch_tech_total
Pivot Table: columns CP:DF
Note: be sure to refresh
table when updating year of
interest on
meet_load_constraint tab
Scenario 1.A: Hourly Dispatch (2035)
Jan Apr
Jul Oct
• VCE is a net buyer during the
early morning hours
throughout the year
• VCE remains being a net seller
but reduces its long position
in the middle of the day with
addition of more storage
• Excess sales unrealistic (see
market constrained scenario)
Dashboard File:
Worktab:
dispatch_tech_total
Pivot Table: columns CP:DF
Note: be sure to refresh
table when updating year of
interest on
meet_load_constraint tab
Scenario 1.A: Market Participation
• In the unconstrained case, VCE
converts from a net buy to a net
seller in 2026 due to favorable unit
economics of candidate projects
reflected in the latest technology
cost assumptions
• Winter months have the greatest
dependency on the market
• Revision of regional modeling
warranted to reflect the impacts of
the IRA on candidate resources
• Dashboard File:
• Worktab: market_participation
• Annual Sales and Purchases: columns B:D
• Annual sales revenue and purchase
expense: columns: columns B:D
Scenario 1.A: Market Participation (cont)
• VCE has open positions in the early
morning hours across the entire
planning horizon
• VCE coverage over the later evening
periods improves over time
• VCE has adequate coverage over the
afternoon peak period across the
entire planning horizon
• Average hourly open positions with
the market differ greatly depending
on the month
• VCE’s long position in the middle of
the day gradually improves over time
• Dashboard File:
• Worktab: market_participation
• Avg hourly net sales and LMPs by
month-year: columns AY:AX
• Avg hourly net sales by year: columns
BE:BK
2030 2035
Scenario 1.A: Market Participation (cont)
• The percentage of hours in the year in which
VCE is short (i.e. purchasing energy from the
market) decreases from 50% and levels off at
around 37%
• The percentage of hours in the year in which
VCE is long (i.e. sells energy into the market)
ranges from 37%-48%
• Although VCE becomes less dependent on
market purchases over time (as reflected by
purchases as a % of annual load), the agency
still gets ~15% of its load from the market in
2035
• VCE sales increase year over year primarily
due to the favorable economics now being
reflected in the IRA-adjusted technology cost
curves, which makes most candidate
resources NPV positive under current LMP
pricing profiles*
• Dashboard File
• Worktab: market participation
• Columns: CM:CP
*Note: First Principles Advisory re-ran the optimization of VCE’s portfolio to reflect the IRA-adjusted cost curves released by the CPUC in early June; these costs curves were not used to determine the PSP developed by the CPUC’s RESOLVE model during the 2022 IRP cycle Scenario 1.A: Costs
• Total costs of the portfolio represent the
aggregate of baseline, candidate and market
• Baseline costs reflect fixed PPA charges for
contracted projects
• Candidate costs reflect forecasted PPA charges from
incremental projects that would be added to the
portfolio
• Market reflects the net costs of load and total
supply settling in the market
• Note congestion is NOT accounted for in this analysis • Total costs are normalized to busbar load and
shown in black (secondary y-axis)
• Net market transactions flips from a net cost to
net benefit in 2026, likely reflecting an
unrealistic assumption that VCE will be able to
secure PPAs and favorable prices relative to
other retail providers
• The impact of this effect on the portfolio costs
increases over time and should be addressed in
subsequent studies
• Dashboard File:
• Worktab: NPV
• Costs information: columns H:N
Scenario 1.A:
Renewable Portfolio Standard (RPS)
• For the baseline case of 80% RPS by
2030, there is no incremental cost in
the model to satisfy this constraint
• The 100% by 2030 constraint in
scenario 1.B also doesn’t bind
• VCE far exceeds its total RPS target in
2030 and beyond
• Additional projects are added to the
portfolio in order to satisfy the RA
constraint
• The model is constrained to ensure
that a minimum of 25% of the RPS
constraint is met with generation
from local projects
• Dashboard File:
• Worktab: RPS
• Costs information: columns P:V
Sensitivity Case: Capped Market Transactions
Sensitivity Case: Market Transactions Policy
• Given VCE’s large concentration in hybrid
projects, the agency will be in a position
to engage in extensive market sales
throughout most of the year
• VCE should define a risk management
policy that defines the maximum allowed
market purchases and sales in any given
hour or slice of day
• Without this risk management policy
defined, future portfolio optimization
exercises will likely result in procurement
decisions that are not in alignment with
prudent scheduling practices
• There is significant variations in costs
depending on what gets assumed in the
model
Mkt sales
uncapped
Mkt sales
capped at 50 MW
Financial Summary: Market Transactions
• When capping market sales to 50 MW in any given hour, the cost premium in increasing VCE’s 2030 RPS target and keeping the 25% local requirement results in an additional charge of $13.5M in present value relative to VCE’s current policy (e.g., 80% RPS with a 25% carve out) • If VCE increases their 2030 RPS target to 100% but doesn’t require any additional local projects results in a cost savings of $19.4M • The NPV of the base case increases by $85.8M in scenario 3, an increase of 13%. This reflects the importance of VCE defining its risk management policy for how it wants to handle its net exposure with the market Index Natural Gas
Price
RPS 2030
Target
Local RPS
Requirement
Market Policy Local Hybrid
Capacity
(MW)
2024-2035 NPV
(2022 $M)
Delta
(2022 $M)
1.A P50 80% 25% Uncapped 65 619.6 0
1.B P50 100% 25% Uncapped 84 643.1 23.5
1.C P50 100% n/a Uncapped 0 586.1 -33.4
3.A P50 80% 25% Capped (50 MW) 65 705.4 0
3.B P50 100% 25% Capped (50 MW) 84 719.0 13.5
3.C P50 100% n/a Capped (50 MW) 0 686.1 -19.4
+$85.8M in costs
when capping sales
Procurement Implications
Procurement Recommendations (redacted)
Technology Breakeven Prices (redacted)
• Breakeven price is the levelized
PPA price that results in an NPV of
zero based on the levelized
benefits of energy and RA*
• Listed prices are the average
breakevens for 2024-2026
• Dashboard File:
• Worktab: Breakeven Prices
• Storage: Column A; rows 123-140
• Non-Storage: Column N; rows 178-214
*Breakeven prices do not include any benefit assumed from PCC1 attributes Potential Next Steps
Topics for Additional Investigation
• Incorporate Slice-of-Day (SOD) into optimization logic used by the model • This study assumed ELCC values, per the methodology utilized in Commission’s IRP cycle • Conduct an all-source solicitation (to collect current market prices for PPAs) and re-optimize portfolio with these costs to determine least cost solution to fill VCE’s near term open position • Optimize the portfolio using nodal prices rather than systemwide zonal prices • This feature will provide a more accurate assessment of curtailment volumes for both baseline and candidate resources
• Direct integration of GHG accounting; 24/7 GHG-free modeling • Integrate CPUC’s Clean System Power GHG accounting methodology (or other logic) into the model to estimate what the incremental costs would be for VCE
• Develop Stochastic Optimization
• This functionality for the handling of uncertainty in key variables within the model when selecting candidate resources
• Currently the model conducts deterministic studies of multiple scenarios to estimate the impacts of uncertainty in future conditions
Key Modeling Inputs / Methodologies
VCE Load
• TEA provided VCE and First Principles Advisory
with an updated net retail forecast across the
entire planning horizon of 2024-2035
• Forecast is NET of BTM solar production
• Updated forecast reflects higher (35% vs 18%)
load growth across the planning horizon of 2024-
2035; this is primarily due to building and
transportation electrification
• Comparatively, the increase in annual peak demand is
18%
• The implicit assumption in the forecast is that
VCE will avoid sharp increases in its peak demand
either passively (from natural customer behavior)
or actively (from customer programs)
• Minimizing increases in peak demand is critical is critical because this impacts the amount of RA VCE will be
responsible for procuring
• The RA constraint is the most binding constraint in the model (i.e., this is what forces VCE to go out and contract for additional resources to be in compliance)
*Note: load in dashboard file represents busbar load (i.e. retail load grossed up for T&D losses) VCE Load Profile
• Customer programs / rate design will be key drivers in shaping VCE’s load shape • Further out on the horizon, winter peak morning leads begin to present similar risks summer peak afternoon loads Jan
Apr
Jul
Oct
RA Forecast (redacted)
NG Profile (redacted)
• Customer programs / rate design will be key drivers in shaping VCE’s load shape • Further out on the horizon, winter peak morning leads begin to present similar risks summer peak afternoon loads Plexos LMP Profiles (P50) (redacted)
• Prices in the late evening / early morning increase
• Even with increasing load, prices in the middle of remain suppressed Jan
Apr
Jul
Oct
Plexos LMP Profiles (P95) (redacted)
• Prices in the middle of the day are elevated in the winter months but muted in remaining parts of the year Jan
Apr
Jul
Oct
Technology Cost Curves
• First Principles Advisory has taken the cost curves from the CPUC’s latest Inputs and Assumptions document which can be found here. • The impacts of the Inflation Reduction Act (IRA) has been applied to all resource types through 2048
• The CPUC applied additional cost modifications to solar, onshore wind, and Li-ion batteries to reflect contemporaneous stressed market conditions
• First Principles Advisory assumed an additional 20% of tax incentives for Li-Ion storage (bringing the ITC up to 50%) for the low-cost battery cost curve
Technology Cost Curves: Overall
• Total levelized fixed costs include
both CAPX and fixed O&M charges
• IRA has increased the cost
competitiveness of clean, firm
resources (e.g., geothermal,
biomass, and non li-ion based
storage)
• RA Only curve represents internal
forecast of annual price for RA only
contracts
Technology Cost Curves: Solar
• IRA enables solar project to now also
have access to the Production Tax
Credit (PTC) which is a tax credit of
$25/MWh
• Given the more favorable economics
of the PTC over the Investment Tax
Credit (ITC), the prevailing notion is
that developers will opt for the PTC
• CPUC assumes developers meet
prevailing wage and apprenticeship
rules to qualify for “Bonus” incentive
Updated
Original
Technology Cost Curves: Wind
• IRA extended availability of PTC,
which was originally scheduled to
expire in 2026 for onshore wind
• CPUC assumes offshore wind will take
ITC; thus no changes from IRA until
post 2035; CPUC also increased the
cost slightly from last IRP cycle
• CPUC assumes developers meet
prevailing wage and apprenticeship
rules to qualify for “Bonus” incentive
Updated
Original
Technology Cost Curves: Storage
• Although stand-alone storage now
qualifies for ITC, the CPUC adjustments
to account for strained market
conditions for li-ion storage result in an
overall cost increase for lithium-based
storage projects
• Flow batteries see a substantive affect
from the IRA tax incentives as does long
duration energy storage (LDES)
technology, which represents iron air
batteries
• CPUC assumes developers meet
prevailing wage and apprenticeship rules
to qualify for “Bonus” incentive
Updated
Original
Technology Cost Curves: Clean, Firm
• Under the IRA, geothermal and
biomass now qualify for the 30%
ITC, previously this was 10%
• CPUC assumes developers meet
prevailing wage and
apprenticeship rules to qualify for
“Bonus” incentive
Updated
Original
ELCC Factors
Modeling Process
Production Cost Modeling
(Plexos)
• IRP Buildout (IRP)
•NG Forecast (TEA)
•GHG Forecast (CEC)
• PCC1 Forecast (TEA)
•Adjusted Net Load / Multiple
Weather Years (FPA/CPUC/CEC)
Portfolio Optimization
(GridPath)
• LMP Profiles (Plexos)
• RA Forecast (TEA)
• PCC1 Forecast (TEA)
• VCE Policy (VCE)
• Step 1: Generate regional WECC prices for 2024, 2026, 2028, 2030, 2032, and 2035 using zonal Plexos model based on CalCCA and First Principles Advisory databases
• Step 2: Along with LMP profiles, apply other market forecasts to VCE portfolio to identify optimal buildout of incremental contracts to determine least cost solution that satisfies the agencies policy goals and regulatory requirements Key Modeling Assumptions
• Typical Week temporal resolution
• 168 continuous hours selected for each calendar month • 2016 dispatch intervals in each calendar year
• 3 adjusted net load weather years modeled in Plexos for each calendar year
• High, med, low
• Average of the 3 runs are applied downstream
• All dollar values are reported in real terms (2022$)
• All time periods are listed in pacific standard time (PST) GridPath vs Match
Differences
Topic VCE PCE
Fundamental Model Plexos WECC zonal
db
Ascend’s
PowerSIMM*
Planning Horizon 2024-2035 2025
Functionality CEM, PCM, PEM PEM
Support /
Development
GridPath is
supported by 3rd
party (Blue Marble)
MATCH is
unsupported
VER profiles n/a Generates profiles
Primary objective
function
Dispatching units in
most economic
manner
Matching hourly
supply to load
RPS Policy Annual 24/7/365
Similarities
Open-source
Month-Hour; Day of
Week Dispatch
Stochastic weather inputs
Deterministic
Scenario analysis