A Battery-Material Plant Is Not a Battery Factory

Sila makes silicon-carbon material for the anode, one of the electrodes inside a rechargeable battery. Its manufacturing roadmap describes capacity in gigawatt-hours, a battery-energy unit that needs an extra set of assumptions before it tells a customer how much material a plant can supply.

The practical lesson is to keep three quantities separate: material produced, battery energy that material could support, and output accepted for a particular customer's cells. A factory can improve one without improving the others. Before treating a large energy figure as deliverable supply, ask for the recipe, production period and acceptance conditions that connect the units.

The product goes inside someone else's cell

Sila's Titan Silicon page, checked October 14, 2026, describes a silicon-carbon anode technology intended to replace graphite and improve cell performance. The company also offers recipe development, cell prototyping and testing with its battery-engineering team.

That second detail helps explain the commercial boundary. A material supplier does not hand over a complete vehicle battery merely by shipping powder. The customer's cell design and manufacturing process determine how the ingredient becomes a working product.

In its December 12, 2023 announcement, Panasonic Energy said it had agreed to purchase Sila material for EV batteries. The release describes silicon's expansion during charging as a challenge and says the partnership would support greater use of silicon in Panasonic's anodes.

That is customer-side evidence of a supply relationship and intended development, not a disclosure of the price, delivered volume or realized performance of every cell. Keep the announcement in its historical period.

Sila's manufacturing page identifies its Moses Lake facility and presents a roadmap of up to 100 GWh of anode capacity by 2030. It separately says shipments from Alameda began in 2021. The roadmap is a future capacity statement, not a verified record of current Moses Lake deliveries.

A large number can conceal a small denominator

A gigawatt-hour equals one million kilowatt-hours. But anode material is a physical ingredient, not stored electrical energy by itself. Expressing its output in battery-energy terms requires an assumption about how much material each unit of battery energy consumes.

Consider a fictional material plant producing 1,000 tonnes a year. Assume an illustrative cell recipe uses 0.10 kilograms of its material per kilowatt-hour. Converting tonnes to kilograms gives one million kilograms. Divide by 0.10 and the output could support ten million kilowatt-hours, or 10 GWh.

Now assume a different recipe uses 0.05 kilograms per kilowatt-hour. The same one million kilograms supports 20 GWh. The plant made no additional material. The energy-equivalent number doubled because the assumed recipe changed.

These are invented inputs, not Sila's recipe, material loading or factory output. They explain why two capacity announcements can be incomparable even when both use GWh.

The denominator also needs a period. A total project aspiration, an annual nameplate rate and the output achieved during a ramp year describe different quantities. A reader should not add “per year” to a source that does not establish it.

Ask whether the capacity assumes a particular cell design and whether it describes rated equipment or actual saleable output. If the supplier cannot share the recipe, it can still describe the basis and bounds of the conversion without disclosing proprietary formulation details.

Yield belongs between production and customer supply

Continue the fictional 1,000-tonne annual example. Assume 85 percent of produced material meets the required specification after ordinary process losses and testing. Accepted output is 850 tonnes, equivalent to 8.5 GWh at the original 0.10-kilogram recipe.

Suppose the plant operates at 60 percent of that assumed annual production level during ramp-up. Saleable output becomes 510 tonnes, equivalent to 5.1 GWh under the same recipe. The sequence matters: 1,000 multiplied by 0.60 and 0.85 gives 510.

This is still a simplified model. Customer acceptance can depend on more than the material's own inspection. If the customer's manufacturing process cannot yet use the material at volume, available supply and consumed supply diverge.

Do not apply the fictional 85 percent or 60 percent to Sila's roadmap. The point is to request separate evidence for utilization and accepted yield, rather than bury both inside a broad capacity figure.

Form Energy's battery-duration comparison addresses another unit problem: energy, power and duration answer different operating questions. A materials plant adds another conversion before any of those battery measures exist.

Build a capacity statement a purchasing team can use

For a supplier evaluation, request a short ledger with four entries:

1. The material output, its physical unit and the period covered.
2. The basis used to convert that output into battery-energy capacity.
3. The specification and accepted yield assumed in the statement.
4. The customer's remaining qualification and production dependencies.

This ledger is an original decision aid, not a statement of Sila's contract terms. It separates a meaningful supplier ambition from an immediately available purchase quantity.

Give every entry an evidence label: observed, contracted, tested under stated conditions or planned. A plan can be commercially useful while remaining a plan. The problem begins when different evidence categories appear in the same forecast without identification.

For a small hardware buyer, the immediate task may be less ambitious than selecting an advanced material. A cell supplier can handle the chemistry while the product team verifies its own dimensions, energy needs and operating conditions. Buying a complete qualified cell is a different transaction from integrating an anode ingredient.

Framework's component-cost explanation shows the value of making a calculation's inputs visible. Here, transparency begins with the unit conversion rather than the selling price.

Keep the ledger with the purchasing forecast so a later revision cannot silently change its basis. If a supplier updates the recipe assumption or the production period, revise the dependent quantities explicitly. A consistent unit is valuable only when the assumptions underneath it stay visible.

Spend the first hour on units

Take one capacity claim from a supplier presentation. Write its unit, time period and evidence status exactly as presented. Then draw the path from that number to the quantity your own production plan needs.

Mark every conversion that lacks an input. It may be material loading, production yield, customer approval or a delivery allocation. Send a specific question about the first missing input before making a purchase forecast from the headline.

Sila's public roadmap leaves material loading, accepted output and complete production economics outside this calculation. Its customer relationship is meaningful evidence, but it cannot reveal the unsuccessful evaluations missing from the public record. Those gaps are reasons to request a better capacity statement before assigning supply to a production schedule.

The defensible conclusion is narrower and useful. A battery-material business must connect manufacturing to a customer's recipe and acceptance process. Until that bridge is explicit, a GWh number describes an assumption as well as a factory.