April 2026 · UC Berkeley · Research Brief

Off-grid RE combined with heat batteries or heat pumps could cost-effectively supply one third of U.S. industrial heat demand.

This dashboard summarizes a facility-level techno-economic analysis of off-grid renewable-powered industrial heat systems. Under the paper's assumptions on technology costs, fuel prices, and local siting potential, modeled opportunities are concentrated in food, chemicals, pulp and paper, refining, and metals.

Based on Dominguez, Paliwal & Phadke (2026)
Facility-level U.S. industrial heat analysis
Data: U.S. Environmental Protection Agency (EPA) Greenhouse Gas Reporting Program (GHGRP) 2022 + U.S. Energy Information Administration (EIA)
FINDING 01 · Cost
1/3
In the modeled 2035 case, off-grid electrification — solar or wind paired with thermal storage and heat pumps — could supply roughly one third of U.S. industrial heat demand at lower delivered cost than natural gas, under the paper's technology-cost and fuel-price assumptions.
FINDING 02 · Siting
5+ yrs
is a representative U.S. transmission interconnection lead time (Lawrence Berkeley National Laboratory [LBNL], 2024). In the modeled behind-the-meter configuration, on-site renewables and storage avoid queue-related delays associated with transmission interconnection.
FINDING 03 · Operations
24/7
Continuous heat service is maintained in the modeled configuration by retaining the existing gas boiler as backup during periods of low renewable output. The analysis therefore evaluates partial fuel switching rather than complete replacement of gas infrastructure.
FINDING 04 · Land
98%
of analyzed facilities have sufficient buildable land within a 15×15-mile buffer for nearby solar to match annual heat demand in the geospatial screening. The remaining ~2% are concentrated in land-constrained industrial corridors and would require alternative supply strategies.
How to read this briefing
What is "industrial heat"?

Industrial heat includes steam, hot water, hot air, and direct process heat used in manufacturing. It accounts for most final energy use in U.S. industry and is still supplied predominantly by fossil fuels, especially natural gas.

Why it matters for the climate.

Industrial process heat contributes roughly ~10% of U.S. greenhouse-gas emissions. The paper examines whether a substantial share of that demand can be shifted to lower-emissions electric heat supply without increasing delivered heat cost.

What does the analysis evaluate?

For each facility, the paper compares delivered heat costs for natural gas, heat pumps, and off-grid renewable systems paired with thermal energy storage (TES), while retaining the existing gas boiler for backup. The siting screen evaluates solar and wind potential within a 15×15-mile buffer around each facility.

What is the policy ask?

The paper implies three practical needs: support for early commercial deployment, faster siting and permitting for industrial-adjacent renewable projects, and financing aligned with 20–25-year industrial asset lives. See § 05 — Policy levers.

§ 01
Site-level cost gap, by scenario and temperature.
Each point represents one industrial facility in the analyzed dataset. Color shows the modeled difference in delivered heat cost between natural gas and the lowest-cost electric alternative at that site, expressed as a percentage of the gas baseline cost for the selected year and temperature range. The CO2-price slider adds a combustion carbon cost to gas only, so the map shows how many facilities change from gas-favored to electric-favored as carbon pricing rises.

Scenario

Temperature Range

Filters

CO2 Price on Gas

0 $/tCO2
Applied to gas only using 0.05306 tCO2/MMBtu and process-specific gas efficiency.

Color Scale: Cost Difference

−75%0%+200%
Percent of adjusted gas baseline: electric option minus gas. Negative values indicate that the electric option is lower cost. Colors are clipped at −75% and +200%.

2030 Scenario · Totals

Sites shown
Sites where electric < adjusted gas
Demand with electric lower cost
Share of demand
Sites flipped by CO2 price
Demand newly below gas
§ 02
Three temperature tiers, three technology pathways.
The lowest-cost electric pathway varies by process temperature. Heat pumps are evaluated only for low-temperature service, abbreviated "LT" (≤200 °C), while thermal energy storage (TES) is evaluated across low (LT), medium (MT, 200–850 °C), and high (HT, >850 °C) temperature ranges. Switching the scenario year shows how relative cost competitiveness changes under the modeled cost assumptions.
Scenario:
Scenario selection updates the cost-competitive demand share in each tier under the current CO2 price applied in Section 01.
LOW TEMPERATURE (LT) · ≤200 °C

Steam, drying, washing

79% of sites · ~7,530 TBtu
Food processing, textiles, low-pressure steam, light chemicals
TBtu cost-competitive
Demand share by cost outcome
Electric option lower cost Gas remains lower cost
MEDIUM TEMPERATURE (MT) · 200–850 °C

Chemicals, pulping, refining

14% of sites · ~1,350 TBtu
Ethane crackers, petrochemicals, nitric acid, pulp & paper digesters
TBtu cost-competitive
Demand share by cost outcome
Electric option lower cost Gas remains lower cost
HIGH TEMPERATURE (HT) · >850 °C

Glass, steel, cement

7% of sites · ~650 TBtu
Glass melting, steel re-heating, cement kilns, hydrogen, lime
TBtu cost-competitive
Demand share by cost outcome
Electric option lower cost Gas remains lower cost
§ 03
Local siting and operational configuration.
The paper evaluates an off-grid configuration consisting of behind-the-meter solar or wind, electric thermal storage, and the existing gas boiler retained for backup. This section summarizes the rationale for that configuration and its operating assumptions.
◉ Off-grid configuration evaluated in the analysis
01
Local renewables
electrons
02 · Path A
Thermal storage
or
02 · Path B
Heat pump (low temperature [LT] only)
heat
03
Plant + gas backup
gas as backup
① Siting
Behind-the-meter renewables

Solar and wind are screened within a 15×15-mile buffer around each facility and modeled as behind-the-meter supply. In this configuration, new transmission interconnection is not required. The median analyzed site has roughly 60× more solar potential than annual heat demand.

② Electrification
Two pathways modeled

Heat pumps are modeled only for low-temperature service (≤200 °C), with coefficient of performance (COP) ≈ 1.9. Electric thermal storage (TES) is evaluated across all temperature ranges; the model assumes 4 hours of storage for low-temperature applications and 15 hours for medium- and high-temperature applications.

③ Backup
Existing gas boiler retained

The modeled system does not retire existing gas boilers. Instead, the gas connection remains available during periods of low renewable output, so reliability is maintained through redundant supply pathways.

Why the paper evaluates an off-grid configuration.

Typical U.S. transmission interconnection wait times have risen to approximately five years, and queued capacity now exceeds the existing installed fleet (Lawrence Berkeley National Laboratory [LBNL], Queued Up 2024). For industrial operators with capital-planning horizons measured in years, that delay constrains deployment of grid-tied renewables.

The paper therefore evaluates an alternative: solar or wind sited within a 15×15-mile buffer of each facility and connected behind the meter. Across the analyzed facilities, the median site has access to roughly 60× more solar generation potential than annual heat demand (first quartile: 20×). Only 73 facilities, or 2.1% of the sample, lack sufficient nearby solar potential for full coverage, but those facilities account for about 27% of total heat demand.

Behind-the-meter siting removes projects from the transmission interconnection queue and avoids grid delivery charges embedded in retail electricity prices. The discussion also notes an operational flexibility benefit: thermal storage can absorb the lower-value fraction of renewable output, potentially reserving higher-value output for grid export where allowed.

Operational continuity through gas backup.

Industrial heat loads require continuous supply. The modeled configuration does not assume retirement of existing gas infrastructure; the gas boiler remains in service when solar and wind output is insufficient. In this framework, solar, wind, and thermal storage provide the primary heat supply, while gas provides dispatchable backup. Reliability is therefore maintained through redundancy rather than full substitution.

How this compares to other decarbonization pathways.

Green hydrogen remains relevant for some very high-temperature processes and applications that require combustion chemistry, but the paper's discussion indicates materially higher delivered heat cost than the electrification pathways shown here under current assumptions.

Carbon capture and storage (CCS) addresses emissions while retaining combustion equipment, but it depends on capture systems, transport infrastructure, and access to storage sites. By contrast, the off-grid electrification pathway evaluated here focuses on reducing delivered heat cost and emissions through local renewable supply and thermal storage.

Grid-tied electrification may deliver similar end uses, but its deployment timeline is affected by transmission availability and interconnection delays. That timing constraint is one reason the paper evaluates an off-grid configuration.

§ 04
Illustrative near-term project pipeline.
This table applies three transparent screens to the selected scenario year: ≥15% modeled cost reduction relative to the adjusted gas baseline, ≥16 megawatts (MW) of thermal demand, and ≥5× nearby buildable land relative to annual demand. Rows are ordered by percentage cost reduction. The list is an analytical screening tool, not a development-ready project pipeline.
Screened projects
Sites that pass all three screens
Annual heat demand
TBtu
across the pipeline
Low-temperature (LT) projects
≤200 °C, food / textiles / steam
Medium-temperature (MT) projects
200–850 °C, chemicals / refining
High-temperature (HT) projects
>850 °C, metals / glass / cement
Filter applied: ≥15% cost reduction vs the adjusted gas baseline in the selected year, ≥16 megawatts (MW) thermal demand, and ≥5× buildable land surplus. This reconstruction reproduces the original embedded facility screen at a $0/tCO2 gas charge and then updates it dynamically as the CO2 price changes.
Click column headers to sort · Scroll for more
Facility State Industry Temp Demand (TBtu/yr) Savings (%) Solar Headroom
§ 05
Illustrative policy levers.
These are illustrative interventions motivated by the barriers identified in the analysis. They are editorial synthesis rather than direct findings of the paper.
Lever I · Federal · Capital

De-risk early commercial projects.

Industrial heat-pump and thermal-storage installations at this scale remain early-commercial. Demonstration grants, cost-share, and concessional debt can reduce first-mover risk while costs and supply chains mature.

DOE Industrial Demonstrations Program · Loan Programs Office (Title XVII §1703 / §1706) · Inflation Reduction Act §48E (clean electricity ITC) · §45X (advanced manufacturing PTC)
Lever II · State / Local · Permits

Unlock industrial-adjacent siting.

The median facility has far more nearby solar potential than annual heat demand within the screened buffer, but a small share of sites remain land-constrained. State and local siting rules therefore materially affect whether modeled potential can be built.

By-right zoning on industrial parcels · Categorical exclusion for behind-the-meter projects · Brownfield reuse incentives · Streamlined interconnection waivers for off-grid configurations
Lever III · Federal · Finance

Match financing to plant life.

Industrial gas assets are long-lived. Standard 7–10-year commercial loans increase annualized cost pressure on capital-intensive clean heat projects, whereas longer-tenor instruments better reflect plant life.

USDA REAP (rural energy) · State green banks · Federal Greenhouse Gas Reduction Fund (EPA) · Private 4(a)(2) green bonds
Lever IV · Demand · Procurement

Pull demand through procurement.

Public procurement can create early markets for lower-carbon industrial products, especially in cement, steel, glass, and chemicals, where buyers can reward lower-emissions production pathways.

Federal Buy Clean Initiative (GSA, DOT, DOD) · State low-carbon material standards (CA, NY, NJ, CO) · EPA EPD requirements for federally funded infrastructure
Caveats and limitations

What this analysis does not claim.

  • Not a full decarbonization pathway. Residual emissions remain because existing gas boilers are retained for backup in the modeled system. Full decarbonization would require complementary solutions such as clean firm power, zero-carbon fuels, or carbon management.
  • Scenario-dependent. The 2030 and 2035 results depend on assumed technology costs, learning rates, fuel prices, and financing parameters. Less favorable assumptions would reduce the share of demand that appears cost-effective.
  • Site-specific risks. Real-world deployment requires site-level engineering, environmental review, labor agreements, and grid-impact studies that the geospatial model does not capture.
  • Not all heat is electrifiable today. Some very high-temperature or flame-contact processes, such as direct-reduced iron and cement clinker production, are not yet direct substitutes for the thermal-storage pathway shown here.
  • Workforce and supply chain. Domestic manufacturing of large-format thermal storage and industrial heat pumps is nascent. Scaling deployment requires parallel investment in installation labor and equipment supply.

Glossary.

Key terms used throughout this dashboard
Behind-the-meter
Generation and storage located on the customer's side of the utility meter. In this dashboard, power is supplied directly to the facility rather than through the bulk grid.
Capacity factor
Fraction of the year a unit operates at full output. A 25% solar capacity factor means annual generation equals 25% of nameplate output sustained over 8,760 hours.
Capital expenditure (CAPEX)
One-time cost of building or installing equipment. Distinct from operating expenditure (OPEX), which is the recurring cost of running it.
Coefficient of performance (COP)
Heat delivered per unit of electricity consumed by a heat pump. A COP of 1.9 means 1 kWh of electricity produces 1.9 kWh of heat.
Industrial heat
Steam, hot air, hot water, and direct flame used for manufacturing processes. Three categories: low (≤200 °C), medium (200–850 °C), high (>850 °C).
Interconnection queue
Backlog of projects seeking permission to connect to the transmission system. Recent U.S. wait times for completed projects are often measured in years.
Levelized cost of heat (LCOH)
Lifetime cost of producing one unit of heat, expressed in $/MWh-thermal. Includes capital, fuel, and operating expenses, discounted to present value.
Levelized cost of electricity (LCOE)
Same concept as LCOH but for electricity. Used here for behind-the-meter solar and wind serving the factory directly.
Thermal energy storage (TES)
Electrically charged materials such as bricks, ceramics, or molten salts that store heat and discharge it on demand. TES acts as a thermal battery that smooths variable renewable supply.
TBtu / MMBtu
Trillion / million British thermal units — units used by U.S. industry to measure heat. 1 MWh-thermal ≈ 3.412 MMBtu.
Solar headroom
Ratio of buildable solar potential within a 15×15-mile buffer to a facility's annual heat demand. The median analyzed site has roughly 60× headroom in the screening exercise.
EPA GHGRP
Environmental Protection Agency Greenhouse Gas Reporting Program. Annual emissions reporting required of large U.S. facilities; the source of facility-level emissions in this analysis.

Interactive levelized cost of heat model.

This panel reproduces the structure of Figure 4A of the paper. Adjust solar or wind capital expenditure (CAPEX), capacity factors, gas price, thermal energy storage (TES) CAPEX, and related parameters to recompute the levelized cost of heat (LCOH) for three representative processes: a steam boiler at 150 °C, an ethane cracker at 850 °C, and a glass melter at 1200 °C.

Use the results to examine sensitivity to input assumptions rather than as a site-specific engineering estimate. Default values reflect the paper's scenario assumptions for the selected year.

About the formulas. LCOH is calculated as annualized capital expenditure (CAPEX) plus operating expenditure (OPEX) plus energy cost, divided by annual delivered heat. For gas, fuel cost enters as gas price adjusted for thermal efficiency; for heat pumps, electricity cost enters as levelized cost of electricity (LCOE) divided by coefficient of performance (COP); for thermal storage, electricity cost enters as LCOE adjusted for the oversizing adder and round-trip efficiency. This simplified reconstruction is checked against Figure 4A of the paper.

1Scenario

2Behind-the-meter electricity

Levelized cost of electricity (LCOE) for solar or wind sited behind the meter at the industrial site. Default values are set to the lower of modeled solar and wind LCOE for the selected scenario year.
$/MWh

3Natural Gas

$/MMBtu
0 $/tCO2
Applied to fuel combustion only; added directly to delivered gas heat cost.

4Thermal Energy Storage (TES)

$/kWh-th
hours
hours
fraction
fraction

5Heat Pump (HP) (low temperature [LT] only)

$/kWt
ratio
fraction
SCENARIO · 2030 · USER INPUTS

Levelized cost of heat under your assumptions

All values are in $/MWh-thermal. Bar heights show modeled LCOH; the outline marks the lowest-cost option for each representative process.

Cost of delivered heat, by process

User-defined assumptions · current-year values
Gas Thermal Energy Storage (TES) Heat Pump (HP) — low temperature (LT) only
Derived electricity price: · Capital recovery factor (CRF): 7% real discount, Gas 20 yr, Thermal Energy Storage (TES) 35 yr, Heat Pump (HP) 20 yr · Gas carbon cost: user-defined, using 0.05306 tCO2/MMBtu and process-specific gas efficiency · TES energy adder: +10% oversizing · HP electricity adder: +$15/MWh at low temperature (LT) for coincidence with storage assumptions (TES has no such adder) · Efficiencies: Gas 80/60/40% · TES 98/95/95% · HP coefficient of performance (COP) 1.9 (low temperature)