Manufacturing Energy Exposure After a Supply Shock: An Evidence Map

An anonymous plant engineer checks utility pipework and a meter beside a manufacturing line.
“The public signal should sharpen the supplier question, not impersonate the supplier's answer.”
— Stan Moskovtsev, Co-Founder & U.S. CEO

Which manufacturing subsectors use the most energy?

Zinit Research charts the energy each manufacturing subsector used per dollar of shipments in EIA’s 2022 survey, and the check under the chart tests the arithmetic behind a supplier’s energy surcharge.

Zinit Research · EIA data, 2022 survey and July 2026 prices

Paper and nonmetallic mineral products used the most energy per dollar of shipments in EIA’s 2022 survey, 3.8 and 3.3 times the manufacturing average

Thousand Btu of energy consumed per dollar of value of shipments, 2022, by manufacturing subsector (NAICS). The survey describes 2022; EIA released these tables between September 2025 and March 2026. Choose a subsector below to highlight it.

Paper9.0
Nonmetallic mineral products7.9
Chemicals4.5
Primary metals4.4
Petroleum and coal products4.1
Wood products2.9
Textile mills2.4
Beverage and tobacco products1.7
Food1.5
Plastics and rubber products1.3
Fabricated metal products1.0
Textile product mills0.9
Furniture and related products0.9
Leather and allied products0.7
Printing and related support0.7
Electrical equipment, appliances, and components0.5
Machinery0.4
Miscellaneous0.4
Apparel0.3
Transportation equipment0.3
Computer and electronic products0.2
All manufacturing2.4

Test a supplier’s energy surcharge

On a 6% energy share, a 15% energy price rise adds 0.9% to the item’s cost, so a 3.0% surcharge is 2.1 points above the arithmetic. Public prices for a plant in Ohio: its average industrial electricity price moved +21.0% from July 2025 to July 2026 and its natural gas price +3.8% from June 2025 to June 2026, weighted by what plastics and rubber products plants spent on each in 2022 (87.8% and 9.9% of their purchased-energy bill), give an energy price change of +19.3%, against the 15% the supplier claims. This is first-order arithmetic before any contract formula; the supplier’s invoices and the agreement decide the number.

Formula: cost effect = energy share of item cost × energy price change, and public price change = (electricity spend × electricity price change + gas spend × gas price change) ÷ (electricity spend + gas spend), with spend from the subsector’s 2022 bill and price changes for the plant’s state. The supplier supplies the share and the price change; the public data only tells you whether they are plausible.

Source: U.S. Energy Information Administration, 2022 Manufacturing Energy Consumption Survey, Table 6.1 (released March 2026) and Table 7.9 (released September 2025), and Electric Power Monthly, Table 5.6.A, released September 24, 2026. Chart: Zinit Research. Data and method

What changed in the public signal?

The October 2 G7 statement says members and partners will implement a coordinated release through the IEA of 100 million barrels over four months, counting commitments already fulfilled, with a substantial diesel release front-loaded within the first 20 days. That is a commitment, with a stated timetable, rather than proof that a specific refinery, market or supplier has received fuel. A buyer can put the announcement on a monitoring calendar. It cannot safely turn the headline barrel figure into an immediate reduction in a supplier's diesel bill or a guaranteed delivery window.

The same statement asks the IEA to monitor implementation and market impact. That distinction gives a practical sequence: log the commitment, look for subsequent implementation evidence, then ask whether the supplier's own procurement and production records show a relevant change. Even a documented market-level release would still leave plant fuel mix, transport routes, inventory and contracts unresolved. Buyers reviewing a price adjustment should distinguish a policy response from the contractual trigger and cost evidence required in their own agreement; the contract escalation guide sets out that separate commercial control.

What does the manufacturing survey actually measure?

EIA released final 2022 MECS results in March 2026. The release describes tables on fuel-consumption ratios, purchased-energy prices and quantities, and building characteristics. The gap between observation year and publication year matters: a 2026 release is not a 2026 energy bill. A category manager can use the tables to learn which kinds of energy questions are meaningful for a subsector or region, with the year and reported unit attached. The tables should not become a current benchmark for an individual supplier's quotation.

The EIA survey report describes an approximately 15,000-establishment sample drawn from a national frame representing roughly 97% of manufacturing payroll. It defines an establishment as one location where manufacturing is performed. This is strong context for sector-level investigation, but an establishment in a statistical survey is not the same thing as a buyer's contracted legal entity or the specific line making its parts. When a supplier operates several plants, the relevant location and process must be named before any energy explanation can be tested. A corporate average may conceal which facility bears the claimed exposure.

The report also distinguishes fuel used for heat, power or electricity from nonfuel energy used as feedstock or raw material. That boundary changes the question the buyer asks. A process using energy as an ingredient can have a different cost mechanism from one using fuel for a furnace or diesel for inbound transport. A quote should identify the affected input, the production step and the evidence of the change; a sector's combined energy total cannot allocate those costs to the buyer's item. Keep the supplier's actual process map next to the public classification rather than assuming the classification is the process map.

The bill is mostly electricity for some subsectors and mostly fuels for others

Electricity was more than half of purchased-energy spending in 16 of the 20 subsectors with a reported total in 2022, so for most suppliers the first document to ask for is the electricity tariff or supply contract. Natural gas carried the largest share in paper, 45.4%, where the gas contract and any hedge matter more. EIA’s survey report also separates energy burned as fuel from energy used as a feedstock or raw material, so ask which of the two a surcharge concerns.

Purchased-energy spending by source, 2022, share of each subsector’s total; the figure at the right is electricity’s share

Computer and electronic products90.8%
Plastics and rubber products87.8%
Apparel86.3%
Printing and related support83.7%
Miscellaneous83.5%
Transportation equipment77.6%
Textile mills77.6%
Machinery77.3%
Electrical equipment, appliances, and components76.6%
Beverage and tobacco products73.9%
Textile product mills72.5%
Fabricated metal products65.3%
Leather and allied products65.1%
Wood products63.6%
Food56.2%
Nonmetallic mineral products50.5%
Paper42.8%
Primary metals40.5%
Chemicals15.4%
Petroleum and coal products14.9%
Furniture and related productswithheld by EIA–

Source: EIA MECS 2022, Table 7.9, expenditures for purchased energy sources. The value shown is electricity’s share. Hatched segments are cells EIA withholds or the remainder they leave; furniture is withheld in full. Chart: Zinit Research. Data and method

Can a historical pass-through estimate settle a current quote?

Ganapati, Shapiro and Walker studied how changes in energy input costs affect producers and consumers in several U.S. manufacturing industries. Their abstract reports a 70% short- to medium-run pass-through estimate for the industries they studied. That finding is about the study's sample and method, not a universal rule. Applying it to a supplier's October 2026 surcharge would cross period, industry, contract and unit boundaries without evidence. It would also confuse an observed market relationship with the allocation of costs under one buyer-supplier agreement.

The study's method accounts for incomplete pass-through, imperfect competition and substitution among inputs. Those considerations are useful prompts for commercial analysis: can the supplier switch fuel, use inventory, change a production route, or absorb part of a cost? They are not answers about this supplier. A buyer should ask for the contract's adjustment formula, the supplier's baseline and current invoices, quantities used for the relevant output, and any offsetting changes. Finance or the category owner can then test the arithmetic and the contract clause; the historical research estimate stays context, not a multiplier.

Which evidence moves the decision?

Public signal to supplier-level test
Public signalQuestion it can frameEvidence required for the buying decision
G7 release commitmentCould fuel availability or logistics change?Implementation readback, relevant supplier fuel and freight contracts, route and delivery records
Historical MECS subsector or regionWhich energy inputs might matter in this process?Named producing site, process energy mix, metered or invoiced consumption for the quoted item
Fuel versus nonfuel distinctionIs energy burned or used as feedstock?Bill of materials and process map, input quantities, conversion and yield records
Historical pass-through researchCould the supplier absorb or reallocate a change?Current contract formula, baseline/current cost evidence, substitution options and quoted margin assumptions

Authored evidence-to-question map, not a supplier score, price model, current benchmark or forecast. Request evidence proportionate to the materiality and decision rights of the specific contract.

Start with one affected item, one producing site and one contract. Write down the claimed change, its date, the source and unit of every number, and the decision it would alter. If the issue is price, test the adjustment clause and a comparable baseline. If the issue is continuity, test alternative energy inputs, capacity and logistics separately; the dual-sourcing strategy guide addresses the latter decision. A supplier can have a genuine cost increase without an immediate continuity problem, or a continuity problem without a supportable surcharge. The worksheet should keep those paths apart.

Where the plant sits changes the energy bill

In July 2026 the average industrial electricity price ran from 4.2 cents per kilowatthour in New Mexico to 39.3 cents in Hawaii, a ratio of 9.5 to one. The U.S. average moved +4.7% from July 2025, and prices rose in 39 of the 51 places with both months reported (the 50 states and the District of Columbia). A surcharge that cites a national rise therefore needs the plant’s state, and better its own invoices, before it can be tested.

Natural gas varies as much. In June 2026 the median state’s average industrial price was $5.88 per thousand cubic feet, and half the states with a reported price fell between $4.37 and $8.43; the extremes were New Mexico at $0.13 and Hawaii at $51.25. The U.S. average moved −6.4% from June 2025. June is the latest month EIA reports in both years for most states; 10 states have no comparable pair and stay blank. For the months ahead, EIA’s Short-Term Energy Outlook of October 6, 2026 forecasts the Henry Hub spot price to average $3.16 per million Btu in 2027, 32 cents below 2026. Henry Hub is a wholesale benchmark and the map shows delivered industrial prices, so a gas surcharge should say which of the two it follows.

Average price of electricity to industrial customers, by state

Cents per kWh, July 2026:under 8.28.2 to 9.29.2 to 10.010.0 to 13.913.9 and over

Highest in July 2026: Hawaii 39.3, District of Columbia 32.1, California 25.1 cents per kWh. Lowest: New Mexico 4.2, Louisiana 6.1, Texas 7.1.

Source: EIA, Electric Power Monthly, Table 5.6.A, data for July 2026 and July 2025, released September 24, 2026; next release October 23, 2026. EIA, natural gas industrial price by state, data for June 2026 and June 2025, released September 30, 2026; next release October 30, 2026. Short-Term Energy Outlook, October 6, 2026. Each square is one state, placed roughly where it sits on a map. Chart: Zinit Research. Data and method

How should the buyer run the next conversation?

  1. Define the claimed exposure: affected part or service, producing site, energy input, relevant contract clause, effective date and requested decision.
  2. Ask for comparable baseline and current evidence: energy invoices or tariff terms, quantities per output, production schedule and any pass-through or hedge terms. Redact unrelated commercial information where appropriate.
  3. Separate price from continuity: test the cost calculation with finance or the category owner; test outage scenarios, qualified alternatives and delivery commitments with operations.
  4. Record unknowns and owners. If the evidence is incomplete, hold the cost conclusion open rather than replacing it with a public-sector average or announced policy outcome.
  5. Revisit after implementation evidence or supplier records change, preserving the earlier source date and decision rationale.

This is not a demand for every supplier to disclose an entire plant ledger. Match evidence to the decision and to what the agreement permits. A routine low-value item may need a short exception review; a constrained process material may justify engineering, operations, finance and legal participation. The buyer can ask the supplier to show a calculation at the level of the affected item while protecting unrelated customer data. If the supplier cannot isolate that item, record the limitation and decide what alternative evidence is adequate, instead of silently treating the broadest figure as precise.

Where should public evidence stop?

Where does this data come from?

What the numbers are
Energy consumption ratios (Table 6.1) and purchased-energy expenditures by source (Table 7.9) for the 21 manufacturing subsectors in EIA’s 2022 Manufacturing Energy Consumption Survey, the average price of electricity to industrial customers by state for July 2026 and July 2025 from the Electric Power Monthly, the average industrial natural gas price by state for June 2026 and June 2025 from EIA’s natural gas price series, and the Henry Hub forecast in EIA’s Short-Term Energy Outlook of October 6, 2026. The survey figures describe 2022 at subsector level; the price figures are state averages across industrial customers. Neither describes a named plant. EIA marks withheld or suppressed cells (Q, D, W or *); they are left blank here.
Source files
How the page stays current
EIA publishes the Electric Power Monthly each month (next release October 23, 2026), the gas price series each month (next release October 30, 2026) and the Short-Term Energy Outlook each month (next release November 10, 2026); the maps are rebuilt from each new release. The survey tables change only when EIA fields the next survey, which it plans for 2027.
Download the data

One ZIP holds the 3 CSV files below and a README that lists the sources and the citation.

  • eia-industrial-electricity-by-state-2026-07.csv
  • eia-industrial-natural-gas-by-state-2026-06.csv
  • eia-mecs-2022-subsectors.csv

Leave your work email and company, and the download starts as soon as you send the form.

We respect your privacy. No spam. Your data is never sold.

Related Zinit Research
Cite this page

Zinit Research. “Manufacturing Energy Exposure After a Supply Shock: An Evidence Map.” Data: U.S. Energy Information Administration, 2022 MECS and Electric Power Monthly. https://zinit.com/journal/research/manufacturing-energy-shock-evidence-map, data retrieved October 8, 2026.

How this page is produced
Zinit Research downloads the source files listed above, keeps the originals, and computes every chart, table and figure in the data sections from them; an automated check rebuilds each number from the files before the page is published. The article text follows the Editorial Policy.

Sources

  1. Manufacturing Energy Consumption Survey (MECS) — U.S. Energy Information Administration, 2026. Historical evidence (official report): Distinguish release date from observation year and available energy fields.
  2. 2022 Manufacturing Energy Consumption Survey Results — U.S. Energy Information Administration, 2026. Historical evidence (official report): Define establishment unit, sample coverage, sector granularity, and fuel/feedstock boundary.
  3. G7 Leaders' Statement on Global Energy Security and Market Stability — G7 Leaders, Presidency of the French Republic, 2026. Contextual evidence (official report): Describe the announced coordinated release and implementation monitoring as commitments only.
  4. Energy Cost Pass-Through in US Manufacturing: Estimates and Implications for Carbon Taxes — Sharat Ganapati; Joseph S. Shapiro; Reed Walker, American Economic Association, 2020. Historical evidence (peer reviewed journal): Explain that even empirical pass-through research is context-specific, not a current supplier benchmark.

Global Procurement Brief

Procurement news, briefed

The market moves, supplier signals, and cost levers that matter — curated by the team behind this Journal. Daily or weekly, your call.

We respect your privacy. No spam. Your data is never sold.

Request a demo
Request a demo