Core economics01What Number Makes or Breaks a Steel Plant?
The decisive number is not the furnace nameplate. It is the contribution earned on every saleable ton at the utilization rate the market will actually support. A steel mill can look profitable at 80% utilization and still destroy cash at 55% because labor, maintenance, environmental systems, rail infrastructure, and debt service do not shrink in proportion to output.
Base-case contribution in the planning model used throughout this article: a $1,025 realized selling price minus $600 per ton of variable metallics, energy, alloys, electrodes, refractories, and variable handling.
That contribution has to cover roughly $7.6 million per month of fixed plant costs in the example 430,000-ton-per-year rebar micro-mill. The resulting operating break-even is about 17,900 shipped tons per month, or close to 50% utilization. Below that line, each idle week becomes expensive twice: revenue disappears, while crews, maintenance capability, environmental monitoring, and financing costs remain.
The U.S. market is large but unforgiving. The USGS 2026 iron and steel summary estimates 82 million tons of U.S. raw steel production in 2025, against about 105 million tons of capacity. It also reports 102 minimills operated by 47 companies. That is enough evidence to reject the comforting idea that installing capacity creates demand.
Do not start with “How many tons can we make?” Start with “Which grades are already committed, at what spread, and how many months of weak utilization can the balance sheet survive?” A signed offtake book is worth more than an optimistic capacity slide.
322,500 shipped tons per year on a 430,000-ton nameplate.
About 17,900 tons per month under the stated cost structure.
Before depreciation, interest, taxes, and owner distributions.
Cuts annual EBITDA by about $32.3 million if scrap costs do not move.
Business model02Which Steel Plant Model Is Actually Financeable?
For a new private developer in the United States, the realistic starting point is normally an electric-arc-furnace minimill or micro-mill, not a greenfield integrated blast-furnace complex. The integrated route needs iron ore, coke production or imported coke, blast furnaces, basic oxygen furnaces, extensive gas handling, and a much larger environmental footprint. It is a national-scale industrial project, not a conventional startup.
The most contained greenfield route: local scrap, EAF, continuous casting, and rolling into rebar or merchant bar. Lower product complexity, but heavily exposed to regional construction demand.
Much larger melt shop and finishing scope. Cold rolling, galvanizing, painting, and grade qualification raise capital intensity but can improve through-cycle margin.
Ore-to-steel infrastructure with long permitting and commissioning risk. Economically sensible only with sovereign-scale capital, established demand, and deep raw-material integration.
The technology choice is also a sourcing decision. EAF economics depend on the quality, chemistry, and delivered cost of scrap and scrap substitutes. Flat-roll or automotive-grade products may require premium scrap, pig iron, direct-reduced iron, tighter residual control, and more finishing investment. A “cheap scrap” strategy can quietly become a quality-rejection strategy.
The U.S. Department of Energy's steel-industry decarbonization review describes the structural difference between the scrap- and DRI-fed EAF route and the ore-and-coke BF-BOF route. That difference flows straight into capital cost, energy exposure, emissions control, working capital, and the kind of customers the mill can serve.
Choose the product first, then the process. Rebar, plate, hot-rolled coil, coated sheet, rail, and special bar quality have different qualification cycles, rolling equipment, scrap chemistry, customer concentration, and margin resilience. “Steel” is not one market.
Startup capital03What Does It Cost to Build a U.S. Steel Plant?
Recent official project announcements show the scale. Nucor's Lexington rebar micro-mill was announced at $350 million for 430,000 annual tons. Its Brandenburg plate mill was a $1.7 billion investment for 1.2 million annual tons. Nucor's West Virginia sheet mill is presented as a $4 billion, 3-million-ton project.
Announced investment per annual ton of capacity
Finishing scope matters as much as furnace size: specialized plate and advanced sheet projects can cost more per annual ton than a simpler rebar or flat-roll configuration.
Calculated from announced investment divided by stated annual capacity. The Sinton comparison uses Steel Dynamics’ original $1.9 billion project estimate and 3.0 million tons of capacity; later reported completed cost was about $2.0 billion.
| Startup category | Focused micro-mill | Large flat-roll mill | What is included |
|---|---|---|---|
| Land, rail, roads, sitework, utility interconnection | $25M | $350M | Acreage, grading, substations, gas, water, rail loop, stormwater, and heavy foundations. |
| Melt shop and continuous caster | $100M | $1.30B | EAF, ladle metallurgy, cranes, scrap charging, caster, oxygen, dust collection, controls. |
| Rolling and finishing | $90M | $1.45B | Rebar rolling at the low end; hot strip, cold roll, galvanizing, paint, and inspection at the high end. |
| Environmental and water systems | $35M | $350M | Baghouse, wastewater treatment, cooling-water systems, monitoring, slag and waste handling. |
| Engineering, construction management, commissioning, contingency | $45M | $350M | Owner's team, EPC support, spares, startup scrap, testing, ramp losses, and contingency. |
| Opening inventory and liquidity reserve | $55M | $200M | Scrap and alloys, payroll, receivables funding, transport deposits, and early operating losses. |
| Total planning range | $350M | $4.00B | Planning range, not a contractor quote. |
The melt shop is not the whole project. Rail, substations, water treatment, foundations, cranes, spare transformers, commissioning consumables, and inventory funding can absorb hundreds of millions. A furnace quote is not a plant budget.
Execution path04How Do You Launch One, and How Long Does It Take?
A credible U.S. launch takes roughly 36–72 months from site control to stable commercial operation. A focused micro-mill may reach the shorter end; a sheet mill with complex finishing, product qualification, and major utility works can run longer. Nucor's Lexington announcement expected two years of construction after permits and regulatory approvals, which is a useful floor rather than a complete development schedule.
Permitting is not a single application. A major source may need a Clean Air Act construction permit and Title V operating permit; water discharges can trigger NPDES requirements; waste handling can implicate RCRA; and state programs add their own siting, stormwater, wetlands, and public-notice steps. EPA's permit-program overview is a useful map, while the actual sequence depends on the state and project design.
Expect $5 million–$25 million for feasibility, front-end engineering, site diligence, air modeling, utility studies, legal work, and owner staffing before the full financing closes.
Carry 6–12 months of fixed cost and startup losses. The line can make steel while still missing yield, chemistry, speed, or customer qualification targets.
A financing-grade launch sequence
- Define the exact product mix, grades, sizes, annual tons, and target customers before selecting equipment.
- Model delivered scrap, DRI, alloy, power, gas, rail, and outbound freight costs by month, not as annual averages.
- Secure site control with permit and utility contingencies rather than buying land outright too early.
- Obtain grid interconnection and natural-gas capacity studies before treating energy cost as a fixed assumption.
- Negotiate offtake, price-index formulas, credit limits, and qualification plans with anchor customers.
- Close equity, senior debt, equipment finance, working-capital lines, and contingency funding as one package.
- Hire operations, maintenance, safety, quality, and procurement leaders early enough to influence design.
Signature economics05Scrap, Electricity, and Yield: The Steel Spread That Pays the Bills
The steel spread starts with realized selling price minus metallics cost, but that is only the first layer. Steel Dynamics explicitly defines metal spread as the difference between average steel selling prices and ferrous scrap consumed. In early 2025, the company reported that selling prices fell faster than scrap cost, compressing steel operating income. That is the pattern a new mill must model: input costs and output prices do not move together or on the same timing.
The $600 variable-cost assumption is a planning case. It should be rebuilt from supplier quotes, delivered scrap grades, power tariffs, yield, and the product route for the selected site.
Recent public-company data provide useful anchors. Steel Dynamics reported $374 per ton of ferrous scrap melted and a $1,107 average external steel selling price in the fourth quarter of 2025. Those figures do not transfer mechanically to a new plant, but they show why a $30–$50 movement in scrap or selling price matters at hundreds of thousands of tons.
Electricity is the next signature line. The EIA reported a 2025 U.S. industrial average of 8.62 cents per kWh. A historical DOE steel-energy study found roughly 770 kWh per shipped ton when the furnace and auxiliary mill loads are combined. At that intensity, every one-cent-per-kWh change shifts cost by about $7.70 per ton, or nearly $2.5 million per year at 322,500 tons.
Impact at 322,500 tons if steel prices and mix are unchanged.
Using 770 kWh per shipped ton as a planning intensity.
Approximate contribution loss at the stated volume and $425 per ton.
The hidden risk is timing. Scrap may reprice monthly while contract steel prices reset with a lag. A mill can show an attractive annual average spread and still suffer a severe quarter because the input and output calendars do not match. Build the model monthly and include contract lags.
Yield and tap-to-tap time complete the picture. Yield determines how much purchased metallic becomes saleable product; tap-to-tap time determines how many heats the asset can produce before downtime and maintenance. The right benchmark is the engineered design for the chosen product, then a conservative ramp below that design. Using mature-mill performance from day one is one of the fastest ways to understate working capital.
Operating cost06What Does It Cost to Run the Mill Each Month?
| Monthly expense | Base case | Per shipped ton | Cost behavior |
|---|---|---|---|
| Scrap, pig iron, DRI, and other metallics | $10.75M | $400 | Variable; quality mix and freight matter as much as posted scrap indexes. |
| Electricity, natural gas, oxygen, and process utilities | $2.02M | $75 | Mostly variable, with demand charges and minimum commitments. |
| Alloys, electrodes, refractories, rolls, and consumables | $1.48M | $55 | Variable by grade, chemistry, campaign life, and process stability. |
| Freight, slag handling, and other variable conversion | $1.88M | $70 | Variable and route-sensitive; outbound freight can decide the market radius. |
| Payroll, benefits, overtime, bonus, and contract labor | $3.00M | $112 | Largely fixed in the short term; shutdowns rarely remove the full labor burden. |
| Maintenance, spares, and planned outages | $2.20M | $82 | Semi-fixed; underfunding turns into longer outages and emergency freight. |
| Insurance, environmental, property, IT, sales, and administration | $2.40M | $89 | Mostly fixed; includes labs, monitoring, security, professional fees, and site overhead. |
| Total monthly operating cost | $23.73M | $883 | Before depreciation, interest, income tax, and owner distributions. |
Labor deserves its own check. Nucor's Lexington project expected about 200 full-time employees. The BLS industry wage profile reported a 2023 mean annual wage of $64,800 across iron and steel mill occupations, with management, engineering, maintenance, and production roles spanning a wide range. A financing model should use loaded cash compensation, overtime, incentive pay, benefits, training, and contractor coverage, not base wage alone.
Protect maintenance and training during the ramp. Cutting those lines can make a monthly budget look better while degrading uptime, yield, safety, and quality—the four metrics that determine whether the plant ever reaches design economics.
Revenue and ownership07How Does a Steel Plant Make Money, and What Can the Owner Take Home?
Revenue is shipped tons multiplied by realized price, adjusted for product mix, contract formulas, extras, freight terms, claims, and rebates. A mill selling commodity rebar has a different revenue profile from a plate or coated-sheet operation that earns premiums for grade, width, coating, certification, or short lead time. The strategic goal is not simply the highest posted price; it is a product and customer mix that preserves spread through a cycle.
Public-company pricing helps anchor the model. Nucor's 2025 Form 10-K reported an average steel-mill sales price of about $1,008 per ton, while Steel Dynamics reported $1,107 per ton in the fourth quarter of 2025. Those are portfolio averages across established assets and product mixes, not a price promise for a new mill. The base case uses $1,025 per ton because it sits inside that observed range.
| Scenario | Utilization | Price/ton | Variable cost/ton | Annual revenue | EBITDA | Cash available to all equity |
|---|---|---|---|---|---|---|
| Conservative | 55% | $950 | $620 | $224.7M | −$14.0M | $0 |
| Base | 75% | $1,025 | $600 | $330.6M | $45.9M | $4.9M |
| Upside | 90% | $1,125 | $610 | $435.4M | $103.3M | $40.3M |
Cash available to equity is deliberately much lower than EBITDA. In the base case, $45.9 million of EBITDA is reduced by an illustrative $22 million of annual principal and interest, $12 million of maintenance capital, and $7 million of tax and reserve funding. The remainder is $4.9 million. A 20% shareholder would receive 20% of any approved distribution, not 20% of revenue or EBITDA.
Base-case monthly bridge from sales to operating cash
A high-revenue plant can still have modest distributable cash because metallics, fixed conversion costs, debt, and sustaining capital are paid before owners.
The owner-income question is therefore inseparable from capital structure. A lightly leveraged plant may distribute more cash but requires more equity upfront. A highly leveraged plant may produce healthy EBITDA and still pay no distributions for years. Management salary should be budgeted as compensation for a job; distributions are a return on equity and should follow lender covenants, reserve policy, and board approval.
Break-even and ramp08Where Is Break-Even, and How Fast Can the Mill Turn Profitable?
At a $1,025 realized price, that equals about 17,880 tons per month, or 214,600 tons per year—approximately 50% of 430,000-ton nameplate capacity.
That is operating break-even before depreciation, interest, tax, and owner distributions. Cash break-even after scheduled debt service is higher. If annual debt service adds $22 million, the project needs roughly another $4.4 million of monthly revenue at the same contribution margin. In practice, a lender will test both the operating line and the debt-service line.
Utilization after first commercial heat
The plant may cross operating break-even around month 10–14, but stable quality, customer qualification, and positive cumulative cash usually take longer.
The market benchmark is a useful reality check. The American Iron and Steel Institute's weekly data reported domestic capability utilization around 79.7% in the week ending July 11, 2026. A startup model that assumes 90% utilization from the first year is not conservative; it is stronger than current industry utilization before allowing for commissioning problems.
Possible after first heat if the plant reaches about 50% utilization and holds its planned spread.
A more realistic window for yield, speed, grade qualification, maintenance routines, and customer acceptance to settle.
For a successful project after commercial operation, depending on cycle timing, leverage, capex, and free cash flow.
Capital stack09How Should the Project Be Funded?
A steel plant is usually funded with a layered capital stack: sponsor equity, strategic-partner equity, senior secured project or corporate debt, equipment financing, working-capital revolvers, state or local incentives, tax-exempt industrial revenue bonds where available, and sometimes federal support for energy or decarbonization features. The financing must cover construction, contingency, interest during construction, commissioning losses, inventory, receivables, and the liquidity reserve—not just equipment invoices.
- 35% sponsor and strategic equity: about $122.5 million.
- 45% senior term debt or project debt: about $157.5 million.
- 10% equipment or vendor finance: about $35 million.
- 10% incentives, subordinated capital, or project grants: about $35 million.
- $35 million–$75 million inventory and receivables revolver.
- Six to twelve months of fixed-cost and commissioning liquidity.
- Dedicated reserve for outage spares, environmental obligations, and debt service.
- Contingency that remains available after mechanical completion.
SBA financing is generally too small for a true steel mill. The SBA 504 program supports long-lived real estate and equipment, but its loan limits are measured in millions, not hundreds of millions. It can be relevant to a downstream processor, fabrication shop, service center, or small rolling operation; it is not the primary capital source for a $350 million greenfield melt shop.
What lenders and investors will demand
- A product-specific market study: customer concentration, delivered-cost radius, import exposure, demand cycle, and replacement risk.
- Firm utility and logistics evidence: power capacity, tariff structure, rail service, gas pressure, water supply, and outbound routing.
- A monthly integrated model: construction draws, capitalized interest, ramp losses, inventory, receivables, debt service, taxes, maintenance capex, and covenant headroom.
- Experienced operators: people who have commissioned and stabilized comparable melt, cast, roll, and finishing equipment.
- Downside liquidity: enough cash to survive lower pricing, 50%–60% utilization, slower qualification, or a major transformer or caster outage.
Do not use every available dollar to complete construction. Mechanical completion without commissioning liquidity can trap the project: the plant is built, but there is not enough cash to buy scrap, carry receivables, replace failed components, or absorb below-plan yield.
Management dashboard10Which KPIs Expose Trouble Before Cash Runs Out?
The useful dashboard is short, operational, and tied directly to cash. Monthly EBITDA arrives too late. Plant leaders need daily and weekly measures that reveal whether tons, spread, yield, energy, uptime, quality, working capital, and safety are moving away from the financing case.
| KPI | Formula | Planning benchmark or warning | Decision it drives |
|---|---|---|---|
| Capability utilization | Shipped tons ÷ rated annual capacity | 70%–85% stable; under 60% is a cash warning in this model | Crew plan, sales urgency, outage timing, covenant forecast. |
| Metal spread | Realized steel price − metallics cost | Base case $625/ton; falling below $550 requires action | Contract repricing, scrap mix, hedging, product mix. |
| Contribution per ton | Price − all variable cost per ton | Base case $425; below $350 raises break-even materially | Order acceptance, surcharge design, mix optimization. |
| Saleable yield | Saleable shipped tons ÷ melted or cast tons | Directional target 92%–96%, product-dependent | Scrap chemistry, process control, cobble and rejection analysis. |
| Energy intensity | Total kWh ÷ shipped tons | Planning range 650–800 kWh/ton across furnace and auxiliaries | Power contract, maintenance, tap practice, schedule. |
| Maintenance cost per ton | Maintenance cash spend ÷ shipped tons | Base case about $82/ton; rising cost with falling uptime is a red flag | Spares, preventive work, contractor strategy, outage scope. |
| Cash conversion cycle | Receivable days + inventory days − payable days | Keep within 45–60 days unless financing is sized for more | Credit limits, scrap terms, inventory levels, revolver need. |
| Recordable injury rate | Recordable cases × 200,000 ÷ hours worked | Track below the industry benchmark; any worsening trend demands intervention | Training, staffing, maintenance access, production pace. |
Safety is a financial KPI, not a separate poster. The BLS industry incidence table reported a 2024 total recordable case rate of 1.9 per 100 full-time workers for iron and steel mills and ferroalloy manufacturing. A serious incident can stop production, damage equipment, trigger investigation, raise insurance cost, and undermine the workforce at the exact moment a ramp needs stability.
- Reforecast cash every week during commissioning, including scrap purchases and customer receipts.
- Review contribution by product and customer, not only average mill margin.
- Separate planned downtime from unplanned downtime and price both in lost contribution.
- Track quality claims and downgrades as dollars per ton, not only tons rejected.
Risk and return11What Can Break the Model, and Is the Investment Worth It?
The project is worth pursuing only when three things are true at the same time: the product has defensible demand inside a competitive freight radius, the plant can source metallics and energy at a durable delivered cost, and the capital structure can survive a weak cycle before the mill reaches mature utilization. Miss any one of those and a technically sound plant can become a financial problem.
| Risk | Trigger | Illustrative annual impact | Mitigation |
|---|---|---|---|
| Steel-price compression | $100/ton lower realized price with no scrap relief | About $32.3M lower EBITDA | Indexed contracts, product mix, low leverage, and downside liquidity. |
| Scrap-cost spike | $50/ton higher metallics cost | About $16.1M lower EBITDA | Multiple scrap basins, owned processing, substitutes, and surcharge design. |
| Utilization shortfall | 10 percentage points below base | About $18.3M contribution loss | Anchor offtake, broader product slate, disciplined commissioning plan. |
| One-week unplanned outage | Transformer, caster, crane, furnace, or rolling failure | About $2.6M lost contribution before repair cost | Critical spares, condition monitoring, OEM support, business interruption coverage. |
| Working-capital squeeze | Receivables and inventory extend by 15 days | Roughly $13M–$18M extra cash need | Borrowing-base revolver, customer credit limits, supplier terms, inventory discipline. |
| Permit or compliance delay | Construction or operating approval slips six months | Interest, owner-team, and standby cost can reach tens of millions | Permit sequencing, early agency engagement, schedule contingency, clear responsibility. |
Environmental rules are part of the economic model. EPA's iron and steel effluent guidelines cover manufacturing, forming, and finishing facilities. Air, water, waste, and monitoring obligations affect both construction cost and recurring operating cost. Treating compliance as an after-the-fact legal line is a budgeting error.
| Payback case | Initial project investment | Annual free cash after maintenance capex | Simple payback | Interpretation |
|---|---|---|---|---|
| Conservative | $350M | Negative | No payback | At 55% utilization, the plant loses EBITDA before debt service. |
| Base | $350M | $34M | 10.3 years | Uses $45.9M EBITDA less about $12M sustaining capital, before financing and tax. |
| Upside | $350M | $80M | 4.4 years | Requires 90% utilization, stronger pricing, and disciplined conversion cost. |
Simple payback ignores the time value of money, taxes, residual value, and financing structure. A full investment case should also calculate net present value, internal rate of return, debt-service coverage, covenant headroom, and downside liquidity.
- A focused EAF micro-mill can be viable, but $350 million is the lower edge of true greenfield steelmaking, not a typical small-business startup budget.
- The biggest value drivers are utilization, metal spread, yield, energy intensity, uptime, and the working-capital cycle.
- Expect two to six years to develop and build, another one to three years to stabilize, and roughly five to twelve years of payback for a successful project.
- Proceed only after a monthly financial model proves the project can survive lower prices, higher scrap, 50%–60% utilization, delayed customer qualifications, and a major outage without running out of cash.
