克里比水资源充足，真正的公用设施缺口在于水处理、污水和危险废物

5.1 A wet coast with a long rainy season

The meteorological station at Kribi has one of the longest records on the Cameroonian coast. The hydrological monograph of the Nyong and coastal rivers, published by ORSTOM in 1979, gives a homogenised mean annual rainfall of 2,835.8 mm for 1937 to 1976, with half of all years between about 2,540 and 3,100 mm [S79]. Modern reanalysis data for the grid cell around Kribi give a similar figure for 1991 to 2020, about 2,836 mm, although reanalysis smooths the extremes [S83]. The coast has an equatorial four-season regime moderated by the ocean: a main dry season from December to February that still brings about 90 to 120 mm a month, first rains from March to June, a short and weak relative dry spell in July, and a main wet season in September and October with about 500 mm a month (Figure 5.1) [S79].

For a steel plant, three consequences follow. Stormwater, not water supply, sets the drainage design. The same monograph estimates a 1-day rainfall of 193 mm with a 10-year return period and 264 mm with a 100-year return period, and records a maximum observed 24-hour fall of 205 mm [S79]. Stockyards for scrap, DRI and slag, which can release suspended solids, oil and metals in run-off, need first-flush retention and treatment sized for these events. Humidity and rain affect materials handling. DRI and HBI re-oxidise and can heat when wet, so covered storage and conveyors are not optional at Kribi. The rain is also a resource, which can be harvested from roofs and paved yards to reduce the plant’s draw on the zone’s supply (below).

5.2 Rivers, groundwater and the limits of local abstraction

Two rivers drain the coastal plateau around the port. The Lobé, which reaches the sea at the Lobé falls north of the port, has a catchment of about 2,305 km² and a mean flow of about 102 m³/s, with a median annual low flow of about 8.4 m³/s and a recorded minimum of 3.5 m³/s [S79]. The Kienké, which reaches the sea at Kribi town, has a mean flow of about 49 m³/s and a low flow of about 6.9 m³/s [S79]. Smaller streams drain the plateau directly to the sea near the port.

In volume, these rivers dwarf any industrial demand: the Lobé’s minimum recorded flow, 3.5 m³/s, is about 110 million m³ a year, fifty times the make-up need of a 1 Mtpa works with closed-loop cooling. Abstraction is nonetheless sensitive. The Lobé falls are on UNESCO’s Tentative List as a cultural landscape [S86], the lower rivers are habitat for species of conservation concern, and the plateau’s small streams, which would receive any treated effluent, have no published flow or quality records. Groundwater is used locally, including by the Kribi gas plant, which is supplied from boreholes [S151], but no public data exist on aquifer yields or quality at the zone.

The sensible approach is therefore a zone-level water system with a single, licensed abstraction designed for the whole platform, rather than a scatter of private intakes and boreholes. That is what PAK has begun to build.

5.3 The port-platform water plant

During his inspection on 19 August 2026, the Minister of Water and Energy visited the works of a dedicated water supply project for the port platform, with an intake and treatment plant whose production is planned for early 2027 [S26]. The plant’s capacity has not been published. Separately, the 20 ha of serviced plots tendered at Mboro in May 2026 include potable-water networks, drainage and a wastewater treatment unit [S18], and KPIZ’s offer to investors includes water and sanitation networks [S03].

These are the right first steps. For heavy industry, three further specifications will matter: the capacity reserved for industrial users and how it grows; whether the zone will supply raw or partially treated industrial water separately from potable water, which is cheaper for cooling make-up; and the tariff. The only published reference for water supplied by the port authority is a 2018 tariff of 400 FCFA per m³ for water delivered to port users [S114]; KPIZ may set its own industrial tariff.

5.4 How much water a steel works needs

Water use in steelmaking varies widely with design. The world steel association’s 2024 sector averages are 8.5 m³ of fresh water withdrawn and 2.3 m³ consumed per tonne of steel [S77]. The difference is returned, usually after treatment, to rivers or the sea. IFC benchmarks for an EAF works are about 3 m³ per tonne for the melt shop, with 1 to 15 m³ per tonne for rolling depending on whether cooling is once-through or recirculated [S75]. A modern closed-loop plant, recirculating more than 95% of its cooling water and treating its scale-pit water for reuse, can hold make-up near 2 m³ per tonne of product, which Cameroon Steel uses as a planning assumption [S112].

Source: worldsteel, Sustainability Indicators 2025 report (2024 data) [S77]; closed-loop make-up target is a Cameroon Steel planning assumption [S112]. Arithmetic in Annex A.

Figure 5.2 shows what the choice of practice means at three sizes. At 1 Mtpa, a plant at the sector-average withdrawal would draw about 8.5 million m³ a year, the equivalent of the domestic consumption of a town of more than 150,000 people at 150 litres per person per day. A closed-loop plant would draw about 2 million m³. The zone should require closed-loop cooling from heavy tenants as a condition of supply, both to protect the shared system and because it simplifies effluent treatment.

5.5 Rain as a resource

Because no month at Kribi is dry, rainwater harvesting is more reliable than in most of the world’s steel regions. A steel works has large roof areas over its melt shop, mills and covered stores, and large paved yards. At the station’s mean rainfall and a run-off coefficient of 0.8, harvesting from 5, 10 and 20 ha of roofs and paved surfaces would yield about 0.11, 0.23 and 0.45 million m³ a year [S79] (Annex A).

Ten hectares of collecting surface would yield about 0.23 million m³ a year, more than a third of the indicative make-up need of a first phase of 0.3 Mtpa [S112]. Storage is the limiting factor, because the wet-season surplus must carry the plant through the drier months, but the cooling-water basins and stormwater retention ponds that a plant needs anyway can serve as storage. Harvested rainwater is also soft and low in chlorides, which suits cooling circuits.

The broader point is that Kribi’s climate rewards integrated water design: the same infrastructure that manages a 264 mm storm can store the water that reduces a plant’s draw on the zone’s supply.

5.6 Wastewater, stormwater and the coast

A steel works produces several distinct water streams: cooling-tower blowdown, scale-pit water from the rolling mill containing oil and suspended solids, water from gas cleaning, sanitary wastewater, and stormwater from stockyards. International lenders apply the IFC effluent guidelines for integrated steel mills, which set limits for suspended solids, oil, metals and other parameters at the point of discharge [S75].

Two local factors raise the stakes. First, the receiving waters are small: the streams that drain the plateau around the port have low dry-season flows and no published baseline. Second, the coast south of the port is protected. The northern limit of the Manyange na Elombo-Campo Marine National Park, created by decree in 2021, lies about 11 km south of the port (Chapter 9) [S84], and the beaches beyond it are used by nesting sea turtles [S91]. Any effluent that reaches the sea near the port must be treated to standards that protect that coast.

For these reasons the zone would benefit from three shared facilities: a central industrial wastewater treatment plant that accepts pre-treated effluent from tenants, a stormwater master plan for the whole zone with retention basins sized for the 100-year storm, and a monitoring network on the receiving streams and the coast. The 20 ha servicing tender’s wastewater unit is a start at plot scale [S18].

5.7 Shared or private utilities: where the line should fall

Every heavy-industry zone has to decide which utilities it provides centrally and which it leaves to tenants. The answer at Kribi follows from scale and risk.

Utilities with strong economies of scale and shared environmental risk should be central. Raw-water abstraction from a single licensed source, a central industrial wastewater plant, a zone stormwater system and a hazardous-waste facility all fall into this category. Each is cheaper per cubic metre or per tonne when built once for the whole zone; each protects the shared rivers and coast; and each is easier to permit once than many times. They are also the kind of public-good infrastructure that the zone’s development-finance package can fund [S07].

Utilities that depend on a tenant’s process should be private, built to zone standards. Cooling towers, scale pits, oil separators, pre-treatment of process effluent and rainwater harvesting belong inside the plant fence, because only the operator can run them with the process. The zone’s role is to set the standards they must meet at the fence: maximum make-up per tonne, effluent quality at the connection to the central plant, and minimum on-site retention for stormwater.

Some services can start private and become shared. A first steel tenant may have to build its own secure cell for furnace dust, or contract an export route, because no shared facility yet exists. If the zone agrees in advance that the cell can later be transferred to, or replaced by, a shared facility, the first tenant’s investment becomes the seed of the zone’s service rather than a stranded asset.

The same logic applies to climate risk. The plateau on which the zone’s industrial plots sit is well above the level of coastal flooding, but intakes, outfalls and any pipelines that cross the shoreline must be designed for a coast that is retreating in places by about 2 m a year and for sea-level rise later in the century [S80][S82]. Designing those crossings once, at zone level, is cheaper and safer than leaving each tenant to engineer its own.

5.8 Hazardous waste: the gap that must be closed first

The most important utility gap for a steel cluster is hazardous waste. An electric arc furnace generates about 10 to 20 kg of dust per tonne of steel, collected in its bag filters (a planning range; the European reference document reports 10 to 30 kg) [S76]. The dust contains zinc and lead, mostly from galvanised and coated scrap, and is classified as hazardous. At 1 Mtpa, one works would produce about 16,000 t a year [S112]. Furnace refractories, oily mill sludge and spent filter media add smaller hazardous streams.

There are three ways to manage furnace dust: recycling in a dedicated zinc-recovery plant, usually a Waelz kiln, which needs a large regional dust supply to be economic; export to such a plant abroad under the Basel and Bamako conventions on transboundary movements of hazardous waste; or stabilisation and disposal in an engineered secure cell. Cameroon’s 2012 decree on waste management sets the conditions for the collection, transport, treatment and final disposal of waste, including hazardous waste [S152]. Our review of public sources did not identify an engineered hazardous-waste facility in the Océan department or a zinc-recovery plant in Central Africa.

For a single first-phase plant, export for recycling or an on-site secure cell is workable. For a cluster, a shared hazardous-waste facility in the zone, licensed and operated to international standards, would lower every tenant’s cost and risk. It also matters to the zone’s reputation: a platform marketed to international lenders cannot have its first heavy tenant stockpiling hazardous dust.

Table 5.1 sourcedUtilities and services a steel platform needs from the Kribi zone: status in October 2026

Source: as listed.