The client calls us to rework a line. After a walk-through we propose building a new one. The client calls us to build a new one. After a walk-through we propose reworking. This doesn't address the budget, it addresses process logic — and usually we're right about one mode versus the original assignment.
The rule isn't in the decision-maker's hands. It's in the state of the current equipment.
Greenfield (new build) makes sense when
The hall's topology doesn't squeeze the new line into bad geometry.
The existing hall may be too narrow, too short, or have columns in the wrong places for the new technology. The layout then works around reality — pour angles, traverse distance, operator access. Every such compromise decision is then paid for by production at every change. For five years.
If the optimal layout requires 80% new concrete, buying a new hall is usually cheaper than reshaping the existing one.
The goal is OEE > 85% from the first quarter.
Brownfield means compromises in services, ventilation, vibration isolation. Often 8–12% of lost OEE compared to the same machine in a properly built hall. If the business case rests on high OEE from the start, brownfield won't deliver.
Technology changes the media — new input material, different waste streams.
Water, gas, air, exhaust. If the new line needs different services than the old one, reworking is more expensive than building from scratch. Installing 80% new services in a hall full of old services is work in a cramped space — slower, more expensive, often more error-prone.
The main investment isn't the line, it's long-term production.
If the business case assumes 10+ years without major changes, the one or two years lost at the start outweighs even a large initial outlay. Brownfield saves now, but occasionally bills you later.
Brownfield (refurbishment) makes sense when
The existing infrastructure is good and the problem is in one machine.
Machine "A" is at end-of-life, machine "B" has reserves. Layout works, services have reserves, people know the environment. Replacing piece "A" with a new one and integrating it into the existing line is always cheaper than rebuilding everything — provided the new machine has compatible interfaces.
Time-to-market is 6 months or less.
A new hall means 12–18 months. Brownfield means 3–6 months. If the business case has a market window, it's often brownfield even when technologically greenfield would be more logical.
The existing team has know-how in the hall you'd lose in a new one.
Underrated factor. In the existing hall people know which sensor is unreliable, which valve needs to be retightened, where the weak spots are. In a new hall you have to discover all this from scratch — sometimes painfully. Brownfield preserves the know-how.
The client has double operations.
Building a new hall next to the old one with production running simultaneously until the new one is proven — this is often brownfield refurbishment after the fact. If operations can't stop, building a new hall is a double solution: expensive and impractical.
Edge cases
There are projects where the answer is "both". Build a new hall for the first part of the process (new technology, new services), keep the existing hall for the second part (storage, dispatch, where a simpler environment suffices). This is hybrid brownfield — more expensive than pure brownfield, cheaper than pure greenfield.
Second case: old hall with new installation, but with the condition that the first year runs in parallel with the old line in another hall. Here it's brownfield + a temporary greenfield investment in a "bridge".
How we decide
A three-hour walk-through. A three-page technical memo. A table with 6–7 criteria (OEE goal, available time, available media, impact on surrounding process, team availability, technology lifespan). If the score comes out +3 or more on one side, it's a clear decision. If it falls between −2 and +2, another hour of detail-hunting.
The worst answer is "let's do a survey". The best answer is a recommendation that holds — or a clear criterion that would change the recommendation.
---
*This decision framework evolved through meetings with dozens of clients in the last three years. No universal formula — but a template that narrows the discussion from "what should we do" to "this vs. that, why".*