Logistics in the FEED phase means the project decides, while the design is still on paper, how the plant will actually be built and delivered: what the largest piece will be, how it will reach site, and which ports, roads and quays it has to pass. Get this wrong and the cost does not show up as a freight invoice — it shows up as a redesign, a split module, or a crane that cannot reach.
Why early logistics involvement in engineering pays
By the time an EPC contract is awarded, the big logistics decisions have already been made — silently, by engineers who were not thinking about transport. Vessel diameter, skid length, the decision to ship a column in one piece or two: each of these is a transport decision disguised as a design decision.
A logistics engineer in the FEED phase costs a fraction of a person-year. A column that turns out to be fifty centimetres too wide for the only available route costs a redesign, a re-fabrication, or a year of waiting for a window. The asymmetry is the whole argument.
Constructability and logistics are the same conversation
Constructability asks whether the plant can be built in the order the schedule assumes. Logistics asks whether the parts can arrive in that order, at that size. These are not two reviews; they are one, and they belong in the same room.
The practical form is a transport envelope issued to engineering early: maximum dimensions and weight per lane, with the limiting factor named. Not “keep it reasonable”, but “4.8 m wide, 5.2 m high, 320 t, limited by the bridge at kilometre 14”. Engineers respect a number with a reason attached.
Fabricators quote per module. Splitting a unit into two smaller modules often lowers the fabrication price and the freight rate at once, which makes it look like an obvious saving. What it adds is site work: a second lift, a field weld, an inspection, a hydrotest, and the scaffolding and crane time around all of it. That cost lands on a different budget line and usually several months later, which is exactly why the trade-off is so often made badly.
What a project logistics plan looks like at FEED stage
It is not yet the logistics execution plan. At FEED stage it is shorter and its job is to constrain, not to instruct:
- Transport envelopes per lane, with the limiting structure or regulation named.
- Candidate routes and ports, with the surveys that still need doing and who pays for them.
- A first pass at modularisation — which units are shipped whole, which are split, and why.
- The long-lead marine items: heavy-lift tonnage availability, quay windows, any specialised vessel that must be reserved far ahead.
- A view on Incoterms, because the contracting strategy follows from it.
Twelve to twenty pages is usually enough. Its value is that it exists before the design is frozen, not that it is complete.
Logistics interface management on an EPC project
Most project logistics failures are interface failures, not transport failures. Procurement places an order against a required-on-site date that logistics never validated. Engineering freezes a dimension that no one checked against the route. Construction plans a lift for a week when the vessel is still at sea.
The fix is unglamorous: one owner for the logistics interface, a standing place on the agenda, and a written record of which assumptions each discipline is working from. On projects with several EPC partners this is the single role that most often goes unfilled, and it is the one I am most often brought in to fill.
Project logistics risk management, concretely
A risk register full of “delay in transport” entries is decoration. Useful entries are specific and have an owner and a date:
- The route survey for lane 3 has not been redone since the roadworks started.
- Only two carriers in the market can lift 900 tonnes at that radius; neither has confirmed availability for Q3.
- The temporary import procedure at the destination has a four-week lead time and nobody has started it.
- The marshalling yard lease ends two months before the last module arrives.
Each of these is cheap to fix at FEED stage and expensive to fix later. That is the test for whether something belongs in the register.
Logistics KPIs on EPC projects
Measure few things, and measure the ones construction actually feels:
- On-site-on-time per lane, not averaged across the project — one bad lane hides in an average.
- Days of float remaining between arrival and required-on-site, trended. Shrinking float is the early warning that nothing else gives you.
- Documentation completeness at departure. Wrong papers stop shipments more often than weather does.
- Cost per tonne delivered per lane, so that a route decision can be argued with a number.
When it is too late
Once the design is frozen and the purchase orders are out, logistics stops being a design input and becomes damage control. Everything is still solvable — there is always a route, a vessel, a permit — but every option now costs money rather than saves it. The FEED phase is the last point at which logistics can make the project cheaper instead of merely possible.