US power-generation economics and utility investment Report Interpretation

Overview

This update refreshes Goldman Sachs’ US generation LCOE model for current construction, operating and fuel costs and post-OBBBA tax-credit scenarios. It concludes that higher costs across technologies raise the capital required to meet load growth, but solar and onshore wind remain among the lowest-cost new-generation options, particularly when safe-harbored, while reliable supply still requires an all-of-the-above mix.

Core views

Goldman Sachs updates its levelized-cost model for new generation using revised construction costs, fuel assumptions and clean-energy tax-credit treatment after the safe-harbor deadline. Costs have increased across most technologies, with particularly large increases in thermal generation: NGCC overnight capital cost rises 40% from $2,000/kW to $2,800/kW; brownfield gas peakers rise from $1,100/kW to $1,400/kW; coal rises from $3,600/kW to $4,500/kW; and advanced nuclear rises from $8,000/kW to $9,000/kW. Utility-scale, commercial and residential solar assumptions rise to $1,250/kW, $1,900/kW and $2,750/kW, respectively. Storage is the exception, declining to $600/kW from $660/kW. The report also raises fixed O&M for several thermal and renewable technologies and introduces a $25/MWh charging-cost assumption for standalone battery storage and solar-plus-storage. Capacity-factor assumptions are largely unchanged, with geothermal and advanced nuclear remaining highest in the analysis. Despite reduced tax credits for projects without safe-harbor status, the report finds utility-scale solar and onshore wind remain economically attractive relative to new gas. In the safe-harbored case, solar and wind LCOEs are about $31/MWh and $27/MWh, respectively, versus about $58/MWh for NGCC. Without safe-harbor treatment, their costs rise to about $47/MWh and $43/MWh, but still remain below gas. The loss of tax credits increases renewable LCOEs by 36%–59% relative to safe-harbored projects, making tax-credit eligibility important for near-term project economics and execution. Goldman Sachs nevertheless expects renewables to help bridge near-term capacity needs because their development timelines are shorter, while gas, storage and nuclear provide dispatchability, flexibility and reliability as load grows. Gas economics are especially sensitive to fuel and construction assumptions. Goldman Sachs lowers its mid-cycle Henry Hub assumption to $3.50/MMBtu from $3.75/MMBtu, but notes meaningful regional price differentials. Under varying fuel-price and capital-cost assumptions, NGCC LCOE ranges from about $37/MWh to $85/MWh, while brownfield gas peaker LCOE ranges from about $80/MWh to $170/MWh. Accordingly, the report argues that utility resource choices depend on location, tax-credit eligibility, fuel prices and construction costs rather than on a single technology ranking. Rising load growth drives the report’s larger industry investment estimate. Goldman Sachs forecasts overall US power demand to grow at a 3.5% CAGR through 2030, including about 108 GW of data-center demand by decade-end. It raises projected utility-scale generation capex for 2026–2030 to about $433 billion from $368 billion, with spending rising across fuel types and NGCC showing the largest capex increase, about 40% versus prior estimates. Generation represents 30%–55% of planned capex across vertically integrated utilities. Solar and storage accounted for about 93% of June net additions, and wind and storage had reached 74% and 63% of Goldman Sachs’ full-year estimates, respectively, reinforcing the importance of safe-harbored renewable and storage pipelines. The GS SUSTAIN analysis distinguishes low-cost intermittent renewable power from round-the-clock reliable power. It sees no US Green Premium for intermittent utility-scale solar and onshore wind, even after incentive sunset, but estimates a Green Reliability Premium when redundancies needed for continuous supply—combinations of solar, wind, batteries and simple-cycle gas, or nuclear—are compared with NGCC. The US base-case premium is $47/MWh at $3.75/MMBtu Henry Hub gas, slightly lower than the prior $48/MWh. Each $1/MMBtu increase in gas price lowers that premium by about $6/MWh, so it is narrower or absent in markets priced off global LNG, such as Europe and Asia. The report emphasizes reliability first, affordability second and emissions minimization third when a premium is involved; near-term US capacity is favored by parts availability and time-to-market for battery storage, solar, wind and simple-cycle gas, followed by NGCC in the medium term and nuclear in the longer term. For equities, Goldman Sachs highlights vertically integrated utilities that can own new projects and earn regulated returns. XEL is Buy-rated and has expanded its five-year capital opportunity to about $70 billion, supported by nearly 13 GW of renewable and battery-storage additions, more than 3 GW of gas through 2030, and more than $10 billion of incremental capex opportunities. DUK is Buy-rated, plans about 15 GW of capacity by 2031, has 7.8 GW of signed data-center service agreements and secured 20 gas turbines. FE is Buy-rated with a roughly 25 GW load pipeline, 6.4 GW contracted demand and a proposed 1.2 GW Maidsville Energy Center. Goldman Sachs is Neutral on AEE, SO and AEP while awaiting greater clarity on incremental generation needs, execution, regulatory outcomes, ownership, financing and contracted-load conversion.

Analysis framework

Goldman Sachs first refreshes technology-specific LCOE inputs for capital costs, O&M, fuel and tax-credit eligibility, then compares safe-harbored and non-safe-harbored generation economics. It tests gas costs against fuel-price and capital-cost sensitivities, translates projected power demand into generation-capex requirements, and assesses how reliability requirements alter the cost of round-the-clock power. It then links potential utility-owned generation investment and rate-base growth to selected covered utilities.

Methodology notes

• Industry AnalysisSupply-demand framework

  US power supply-and-demand and generation-capex model

  The report uses projected load growth, including data-center demand, to estimate the new generation capacity and capital spending required through 2030.

• Other

  Levelized cost of energy (LCOE) analysis

  The model compares the lifetime per-MWh cost of new generation technologies using capital costs, operating costs, fuel costs and tax-credit scenarios.

• Valuation methodsP/E and PEG Valuation

  P/E price-target methodology for XEL, DUK, FE, AEE and SO

  The report’s price-target table applies specified forward-EPS P/E multiples to these covered utilities.

• Valuation methodsSOTP (Sum-of-the-Parts) Valuation

  Sum-of-the-parts price-target methodology for AEP

  The report identifies SOTP as the target-price framework used for AEP.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Xcel Energy (XEL)

  Buy-rated beneficiary of utility-owned renewable, storage and gas investment and data-center-load conversion.

  Strengths

  ~$70 bn five-year capital opportunity; nearly 13 GW of renewables and storage, more than 3 GW of gas through 2030, and more than $10 bn of incremental capex opportunities.

  Comparison

  Highlighted as particularly well positioned through renewable scale and selected company-owned RFP projects.

  Risks

  Negative rate-case outcomes, litigation, failure to close the earned-versus-authorized ROE gap and cost management.

• Duke Energy (DUK)

  Buy-rated beneficiary of growing large-load demand and an all-of-the-above capacity plan.

  Strengths

  ~15 GW planned capacity by 2031, 7.8 GW of signed data-center service agreements and 20 secured gas turbines.

  Comparison

  Combines gas under construction with solar, storage, nuclear uprates and advanced-nuclear opportunities.

  Risks

  Balance-sheet improvement, regulatory uncertainty and a decline in the load-growth forecast.

• FirstEnergy (FE)

  Buy-rated potential beneficiary of generation and transmission investment tied to accelerating data-center demand.

  Strengths

  ~25 GW total pipeline, 6.4 GW contracted demand and a proposed 1.2 GW Maidsville Energy Center.

  Weaknesses

  Opportunity requires regulatory approvals and customer-cost protections.

  Risks

  An unfavorable Ohio rate-case outcome, higher pension or interest expense, and mild weather weighing on earnings.

• Ameren (AEE)

  Neutral-rated utility with growing Missouri large-load demand and an expanding generation plan.

  Strengths

  2.8 GW of signed ESAs; ~2.25 GW of approved, under-construction or planned solar, storage and gas capacity for service by 2027/28.

  Weaknesses

  Goldman Sachs awaits greater clarity on incremental generation needs and related capex.

  Risks

  Load growth, regulatory outcomes and macroeconomic conditions affecting data-center deals.

• Southern Co. (SO)

  Neutral-rated utility with a large contracted-load pipeline and company-owned generation buildout.

  Strengths

  More than 17 GW contracted large-load pipeline through the mid-2030s and 10 GW of company-owned thermal, solar and battery generation.

  Weaknesses

  Execution and regulatory considerations support the Neutral rating.

  Risks

  Negative regulatory developments, revisions to load growth and balance-sheet considerations.

• American Electric Power (AEP)

  Neutral-rated utility with potential incremental generation and transmission investment from contracted load.

  Strengths

  69 GW contracted-load pipeline through 2030, 13 GW of secured gas-turbine capacity and evaluation of an additional 10 GW through 2035.

  Weaknesses

  Project timing, ownership, financing and load conversion remain uncertain.

  Risks

  Higher-than-expected financing needs, failure to close the ROE gap, cost management and uncertain conversion of contracted load into capital investment.

Key data

• Utility-scale generation capex through 2030~$433 bnRaised from the prior ~$368 bn estimate.
• US power-demand growth3.5% CAGR through 2030Includes approximately 108 GW of data-center demand by the end of the decade.
• Safe-harbored utility-scale solar LCOE~$31/MWhVersus ~$58/MWh for NGCC.
• Safe-harbored onshore wind LCOE~$27/MWhVersus ~$58/MWh for NGCC.
• Non-safe-harbored solar and wind LCOE~$47/MWh and ~$43/MWhBoth remain below the report’s ~$58/MWh NGCC estimate.
• NGCC overnight capital cost$2,800/kWUp 40% from $2,000/kW.
• US Green Reliability Premium$47/MWhAt $3.75/MMBtu Henry Hub gas, versus $48/MWh previously.
• Global data-center power demand1,506 TWh in 2030EUp 1,107 TWh from 399 TWh in 2023.

Impact & implications

Goldman Sachs says the combination of higher construction costs and rising load expands the US generation-capex opportunity, particularly for regulated utilities able to own projects. Renewable and storage additions remain important for near-term capacity, but the report expects gas and nuclear to remain necessary for reliable power as demand grows; the economic mix will vary with regional gas prices, project costs and tax-credit eligibility.

Risks

• Renewable LCOEs rise 36%–59% without safe-harbor tax-credit eligibility.
• Gas generation economics vary materially with regional fuel prices and construction costs.
• Project economics depend on location, tax-credit eligibility, fuel prices and construction costs.
• Utility opportunities depend on execution, regulatory approvals, financing, affordability and conversion of large-load pipelines into firm capital investment.

What to watch

• Whether data-center demand converts into firm utility commitments and incremental generation or transmission capex.
• Safe-harbor tax-credit eligibility across renewable and storage pipelines.
• NGCC construction costs, regional gas prices and resulting LCOE sensitivity.
• Regulatory approvals and customer-cost protections for proposed utility generation projects.
• Availability of turbines, other parts and labor, which affects the timing of capacity additions.