A battery pack assembly plant can look simple on paper: buy cells, assemble them, sell packs. The margin sits in the difference between the value of the finished product and the cost of cells, other materials, conversion and overhead. But a business case built on one spreadsheet row per year rarely survives the first meeting with a credit committee.
A bankable model shows how each number is built, how it moves when assumptions change, and how the plant behaves during the difficult first two years.
1. Start from the line, not from the market
Market forecasts tell you whether there is demand. They do not tell you what the plant can produce. Capacity should be derived bottom-up: the bottleneck cycle time of the line, the shift pattern, the working days and a realistic overall equipment effectiveness (OEE). If the capacity in the model cannot be traced to a line design, a reviewer will discount it.
2. Separate the cost blocks
| Block | Includes | Main drivers |
|---|---|---|
| Initial investment | Equipment, conveyors and AGVs, MES and IT, installation, commissioning, building fit-out | Automation level, capacity, product complexity |
| Materials | Cells, busbars, CCS, BMS, enclosures, thermal materials, packaging | Cell price, design, scrap rate |
| Conversion | Direct labour, energy, consumables, maintenance | Labour rates, automation, OEE |
| Overhead | Quality, engineering, logistics, management, certification | Plant size, product mix |
Materials, and cells in particular, usually dominate the cost of a finished pack. That makes the model highly sensitive to cell purchasing terms and to scrap, and less sensitive than many expect to the equipment price itself. Model them explicitly.
3. Model the ramp-up honestly
No line runs at nameplate capacity in its first month. Plan a ramp-up curve with yield and OEE improving over several quarters, and fund the working capital it consumes: cells bought before packs are sold, scrap during process validation, and overtime during qualification builds.
4. Show the automation trade-off
Automation moves cost from operating expense to capital. A good model shows at least two configurations side by side, for example a semi-automated line and a fully automated line, with the investment, headcount, cost per kWh and payback of each. This lets decision makers choose the risk profile they prefer rather than accept a single answer.
5. Stress the assumptions that matter
Run sensitivities on the variables that move the result most:
- Utilisation: what if volumes reach only 60% of plan in year two?
- Cell price and payment terms
- Labour rates and availability per shift
- Yield and scrap during ramp-up
- Incentives and local-content premiums, and what happens if they change
Reviewers do not expect your forecast to be right. They expect your model to show what happens when it is wrong.
6. Make it auditable
Keep inputs, calculations and outputs on separate sheets, document every source and avoid hard-coded numbers in formulas. A clear model shortens due diligence and builds confidence in the team behind it.
Try it on your own line
Our 3D line configurator lets you switch each workstation between automated, semi-automated and manual operation and see investment, capacity and headcount change live.