Terminal advisory
A successful bulk terminal depends on every stage working together: vessel access, unloading, transfer, storage and dispatch.
InVenture brings commercial and technical expertise together to support terminal design and implementation. We consider the cargo, the site and the intended operation as one system.
Two sections through the same trade, drawn to scale. Alongside, the depth you have decides the class you can work and the parcel that follows. When the water is not there, the cargo still moves — at anchorage, into lighters, through a floating crane, and at a price paid in cycle time. Change the vessel, the depth and the tide, and the plate redraws.
Drawn to a single scale, horizontal and vertical alike — no exaggeration. Vessel particulars are indicative class figures for planning, not any particular ship, and the tidal levels are an example regime until survey establishes yours. Real work starts from the chart datum on your own survey, the controlling depth along the approach — usually shallower than the berth — and the squat and motion allowance your under-keel policy is written against.
Throughput is decided long before the cargo arrives. The depth carried along the approach and alongside sets the vessel you can take; the vessel sets the parcel; the parcel sets what the terminal can move in a year. Get that chain wrong at design stage and no amount of equipment fixes it.
Indicative only. Real figures come from survey: declared depth against chart datum, tidal window, under-keel clearance, and the controlling depth along the approach — which is often shallower than the berth itself.
The controlling depth between the roads and the berth, tidal windows, and how long a vessel waits before it can come alongside.
Depth alongside against chart datum, under-keel clearance, and the largest vessel the berth can actually work rather than nominally accept.
Cranes, grabs, conveyors and sheds sized to the grade — and the annual tonnage that combination genuinely supports.
Carrying a terminal from layout and equipment selection through to a facility that runs at the rate it was drawn for.
A berth run at 90% looks like an asset being used properly. It is usually a queue. Waiting time does not rise with occupancy in a straight line — it accelerates, and it accelerates hardest where there is only one berth. Move the sliders and watch where the curve leaves the floor.
Screening tool, not a design basis. Computed live from the Erlang-C formula for an M/M/n queue — random arrivals, exponentially distributed time alongside, first come first served, steady state, all berths able to work all calls. Our implementation reproduces the published UNCTAD queueing table at all twelve points we test it against, including a control that fails deliberately if the check itself stops working. Real arrivals are rarely Poisson: a contract trade with nominated arrival windows queues considerably less than this curve shows, spot and seasonal trades rather more. It models none of what actually stops a berth — weather, tidal window, draft restriction, cargo changeover, equipment failure, stockyard saturation, working-hour rules. Use it to see the shape of the risk, not to size a facility.
Occupancy tells you a ship was there. It does not say whether anything was being done. Split the occupied time four ways and the answer usually changes the investment case — because the first reaction to a congested berth is to build another one, and the cause is frequently that the berth already there is worked for a third of the time it is occupied.
An illustrative case at 85% overall occupancy, not a specific facility. Recover the two unproductive bands and the same quay carries the same tonnage with the queue collapsing — a far cheaper answer than a new berth, and the one that is routinely skipped because occupancy alone never showed the problem. Note also that the recording interval alone moves the headline: tick a box once a day and a week that measured 70% by the hour reports as 90%.
The two are not the same measurement and they do not move together. You can raise output at any time by adding people and machines. Whether that was worth doing is a different sum, and it is the one that decides an equipment case.
| Illustrative berth operation | Present | Heavier crewing |
|---|---|---|
| Workers per gang | 20 | 26 |
| Cranes per gang | 1.0 | 1.5 |
| Gangs working the ship | 2 | 3 |
| Output per gang-hour (t) | 20 | 25 |
| Output per ship-hour (t) | 40 | 75 |
| Crane cost per ship-hour | $200 | $450 |
| Labour cost per ship-hour | $240 | $468 |
| Handling cost per tonne | $11.00 | $12.24 |
Output per ship-hour rises by 88%; the cost of handling a tonne rises by 11%. The heavier option only pays if the time saved is genuinely realised somewhere — and it evaporates the moment the vessel that finished early then waits on a pilot, a tide or a berth. That is why we model the terminal as one system rather than optimising the quay in isolation. Illustrative figures at round-number rates, shown for the arithmetic, not as market prices.
Quay equipment is chosen from a catalogue. The stockyard is the decision you cannot revisit: layout fixes capacity, blending ability and which grades you can hold apart, for the life of the site. A yard with no room to keep two qualities separate cannot buy the cheaper cargo and blend it — the commercial option is lost at the drawing stage, years before anyone notices.
Suits long sites and very large tonnages, into the millions. Capacity is extended cheaply later by lengthening the track and the conveyors, and long-term and short-term piles can sit side by side.
Extendable · high capacity · longer civilsSuits squarer sites to roughly a hundred thousand tonnes. Lower capital and operating cost, but capacity is fixed once built and the shorter layer length means input lots must be smaller to blend properly.
Compact · fixed capacity · small lotsStacked continuously on one axis. Simple and cheap; poor blending, and the coarse fraction rolls to the toe.
Segregation high · blending poorRoof-shaped layers from one discharge point on the centreline; reclaimed from the end, across every layer. Good blending, still segregates by size.
One discharge point · end reclaimRows laid by a slewing boom with several discharge points. Costlier stacker; segregation minimised, which is what a wide size range needs.
Multi-point · least segregationCone-and-chevron on a circular bed around a central column: stacking the leading face while reclaiming blended material behind it. Continuous.
Circular · stack and reclaim togetherBlending improves with the number of layers, and every linear bed has the same known weakness: the end cones do not contain all the layers, so the first and last of a pile run off-grade unless they are recycled. That is predictable at design stage and can be quantified in advance — which is the difference between a grade excursion you planned for and a claim.
Side-working bucket wheels are selective — they will take a hot or off-spec area out of a pile — and reach very high rates, but they never see the whole cross-section, so they blend less well. Bridge and portal scrapers work the full face and blend excellently; they cannot step over a neighbouring pile. Choose for the property you actually need, then lay the yard out around that machine.
Selectivity and blending pull opposite ways.Free surface is what the wind erodes. One heap holding what two held cuts exposed surface by about a quarter; the long axis wants to lie along the prevailing wind; an upwind retaining wall or embankment removes more surface still, and windbreak netting has been measured at roughly halving wind velocity. Each of those has a price — walls restrict access, netting has a short life and traps birds, and moving the pile shape moves your capacity.
Shape the pile before you buy the sprays.Minimising wind-exposed surface wants steep, tall, walled piles. Controlling self-heating in a reactive material wants shallow slopes, dense compaction and no stagnant material against a wall. Both are correct and they cannot both be satisfied. Where a cargo can heat, the shallow answer wins and the lost capacity is designed for — not discovered later.
Safety-critical direction wins; capacity is the price.Grids of buried thermocouples suit a strategic pile, not one being stacked and reclaimed daily — and early hot spots are local enough to sit between the probes. Working yards rely on rapid turnover, compaction discipline, gas indication and, above all, on not stacking incompatible material together in the first place. Detection is not a substitute for layout.
The monitoring plan starts at the layout.
Terminal plant comes from four separate supplier groups with different economics. The nameplate on a thirty-year-old machine is often not the company that can support it today — several of the great European lineages have changed hands.
Engineered one-offs. Specialists cluster by lifting mechanism — screw, bucket chain, pneumatic.
Bruks Siwertell · NEUERO · VIGAN · Bühler · ZPMC · DHHI · IHI · Kawasaki · SumitomoCatalogue machines with options — and the grab bought separately, from a separate specialist.
Konecranes · Liebherr · Sennebogen · Mantsinen · Grabs: Nemag · Verstegen · ORTSStackers, reclaimers, blending beds, overland and pipe conveyors — sold as systems.
KOCH Solutions · TAKRAF · BEUMER · BedeschiThe lower-capital tier, for terminals that cannot yet justify a fixed machine.
Telestack · SAMSONA rated capacity in a brochure is a claim. What a terminal gets across a full vessel depends on hold geometry, clean-up and the plan agreed with the ship.
Every terminal has one constraint that is actually binding, and it is rarely the one being discussed. Our engagement is built to find it, put a number on it, and say what changes — in that order, and in writing.
What actually crosses the quay, by grade and not by trade name: size distribution, moisture, density, dust class, abrasion, corrosion, heating behaviour, and how that mix varies through a year.
Output — cargo basis with the handling implications statedControlling depth on the approach against chart datum, tidal window, under-keel clearance, and the largest parcel the berth can work rather than nominally accept.
Output — admissible vessel and parcel envelopeOccupancy split into vacant, occupied-not-working, occupied-and-working and occupied-not-workable, with queueing exposure modelled against it. This is where a proposed new berth usually turns into a shift pattern.
Output — occupancy, waiting exposure, recoverable hoursRated against achieved capacity for each leg — ship-to-shore, transfer, stacking, storage, reclaim, delivery out — to find the leg that is setting the terminal's real rate.
Output — bottleneck identified and quantifiedStorage capacity against the trade's stock cycle, blending requirement, segregation of grades, dust and run-off control, and the conflicts between those objectives resolved explicitly rather than averaged.
Output — yard layout and control strategyThe annual tonnage the facility genuinely supports today, the tonnage each intervention buys, what each costs, and the order to do them in — including the interventions that cost nothing but a decision.
Output — staged plan, costed, with the capacity each stage releasesSpecification, procurement support and site presence through commissioning, so that the facility is proved against the throughput it was designed to — not handed over against a drawing.
Output — a terminal running at the rate it was drawn forEngagements are scoped to the facility. Nothing on this page is a design basis, a compliance opinion or a substitute for site-specific engineering, permitting and statutory sign-off.
For a cargo enquiry, tell us the grade, the tonnage and where it has to be delivered. For a terminal, the useful first message is shorter than most people expect: the depth you have, the cargo you handle and the tonnage you are trying to reach. That is enough to say whether the constraint is the water, the quay or the yard — and we will say which before anything is scoped.