Last updated: November 2026 · Reading time: about 11 minutes
 |
| The pipeline replaces the port-to-border truck leg and one transhipment at Dewele. The inland run from the border to Addis remains by road. $660m removes one handover, not the whole chain. |
What this guide covers
What was announced
The project is a 120-kilometre refined product pipeline running from Damerjog in Djibouti to Dewele on the Ethiopian border, at a reported cost of $660 million. It is the smallest of the three assets in Dangote’s announced African distribution network — the Walvis Bay terminal is priced at $2.5 billion, and the Kenya Pipeline stake at $327.4 million — and it covers the shortest distance of any of them.
Damerjog is a newer port development on Djibouti’s coastline, distinct from the older Doraleh facility that has handled most of Ethiopia’s trade. Dewele is the road crossing on the Ethiopian side, already an established trucking entry point.
The pipeline replaces a segment that currently moves by truck: from the Djibouti port to the border, where product is then loaded onto Ethiopian-registered trucks for the inland run. Replacing that segment with pipe removes a transhipment, a border queue and a fleet of tankers from the chain.
The $5.5m/km question
Divide the cost by the distance and you get about $5.5 million per kilometre. A typical onshore trunk line in reasonable terrain costs materially less than that. So either the project is unusually expensive, or the $660 million figure is buying more than pipe.
The second explanation is almost certainly the right one, and it is inferable from what a functional pipeline at Damerjog would require:
- Port and marine works. A refined product pipeline has to originate somewhere a tanker can discharge. If Damerjog is a new development rather than an existing terminal, the berth, the discharge line and the marine works are inside the project cost.
- Storage at the origin. A pipeline cannot run directly from a ship. It needs buffer tanks at the port, sized to match discharge rates and pipeline throughput.
- Border and customs infrastructure at Dewele. Product crossing a frontier needs metering, documentation and a physical facility for the transition. That is a building and a system, not a length of pipe.
- Loading facilities on the Ethiopian side. Because the pipeline stops at the border, it needs a terminal there to load product onto Ethiopian trucks for the inland run.
- Terrain and security. The Djibouti corridor runs through arid, sparsely populated terrain, and the security environment along the route is a cost line that does not appear on an engineering drawing.
That matters because it changes the comparison. A $660 million project that includes port works, storage, a border facility and 120km of pipe is a different proposition from a $660 million pipeline. The first is a piece of national infrastructure. The second would be overpriced.
Ethiopia’s chokepoint
Ethiopia is landlocked. It has no coastline, no refinery of meaningful scale, and no domestic crude production. Every litre of refined product consumed in the country arrives by ship at a foreign port and then travels inland by road or rail.
For decades, that port has overwhelmingly been Djibouti. The Djibouti corridor carries the large majority of Ethiopian trade — by most estimates well above 90% — and it is the route for the Addis-Djibouti railway as well as the road network. That concentration is the chokepoint, and it is a genuine strategic vulnerability for a country of more than 130 million people.
The vulnerability is not theoretical. It has three dimensions:
| Dimension | What it costs Ethiopia |
|---|---|
| Single-route dependence | Any disruption at the port — congestion, industrial action, a geopolitical dispute with Djibouti — immediately affects fuel supply to the entire country. |
| Transhipment and trucking cost | Product is discharged at the port, loaded onto trucks, driven to the border, cleared, transhipped onto Ethiopian trucks, and then driven inland. Every handover adds cost and time. |
| Border friction | The Dewele crossing is a queue. Customs processing, documentation and vehicle inspection add days to a journey that is physically short. |
A pipeline from Damerjog to Dewele addresses the second and third of these directly. It replaces the port-to-border truck leg with continuous flow, removes one transhipment entirely, and takes a fleet of tankers off a road that is also carrying general freight.
It does not address the first. The pipeline still starts in Djibouti. If the Djibouti relationship deteriorates, the pipeline is as exposed as the road was — arguably more so, because a pipeline is a fixed asset that cannot be rerouted, whereas trucks can be sent to another port.
The diversification tension
This is the part of the project that generates the most interesting question, and it is barely discussed in the coverage.
Ethiopia has been actively trying to reduce its dependence on Djibouti. That policy is not new and it is not secret. The government has pursued alternative corridors — Berbera in Somaliland, Port Sudan in Sudan, and Lamu in Kenya — for exactly the reason described above: a single-route dependence on a neighbour is a strategic risk that a landlocked country cannot accept indefinitely.
A pipeline that runs from a Djibouti port to the Ethiopian border reinforces the route Ethiopia is trying to diversify away from. It makes the Djibouti corridor more efficient, which makes it more attractive, which makes it harder to shift volume to an alternative.
There is a reasonable answer, and it is worth stating because it is the strongest version of the case. Ethiopia’s diversification efforts have been slow and have not yet produced a corridor that can carry meaningful volume. Berbera requires road and rail investment on the Ethiopian side that does not exist at scale. Port Sudan has its own instability. Lamu is a long way from being operational as a fuel route. In that context, making the existing corridor more efficient is not a substitute for diversification — it is what a country does while diversification is being built.
But that argument only holds if the pipeline is genuinely cheaper per litre than the truck route it replaces. If it is, the efficiency is real and Ethiopia benefits immediately. If it is not — if the pipeline tariff is set at a level that recovers the $660 million over a reasonable period — then Ethiopia has paid for a fixed asset that makes its dependence more expensive rather than less.
No tariff has been published. That is the number that determines which of these it is.
Why the smallest asset matters most
Place the three assets side by side and the logic becomes clearer.
| Asset | Cost | What it does for the network | What it does for the market |
|---|---|---|---|
| Walvis Bay terminal | $2.5bn announced | Creates a storage and distribution node on the Atlantic coast | Adds capacity to a market that already has alternatives through Durban and Beira |
| Kenya Pipeline stake | $327.4m | Buys access to an existing inland network | Positions the group inside a route that already works |
| Ethiopia–Djibouti pipeline | $660m | Removes a transhipment bottleneck on a single-route corridor | Creates a route where none existed |
The distinction in the last column is the argument. Walvis Bay adds capacity to a market with competition. The Kenya stake buys into a route that already works. The Ethiopia pipeline is the only one that changes the physical structure of supply for a market that has no alternative.
That is why 120km can matter more than 1,700km. The Namibia terminal will succeed or fail on whether it can win volume from incumbents. The Ethiopian pipeline succeeds simply by existing, because the alternative is a truck queue at a border.
Who pays for it
The financing structure has not been disclosed, and that is the most important gap in the announcement after the tariff.
The project sits at the intersection of three financing questions:
- Who owns the asset? A cross-border pipeline requires an ownership vehicle that both Djibouti and Ethiopia accept. That may be a joint venture, a special purpose company, or a concession. Each has different implications for tariff setting and regulatory oversight.
- Who carries the construction risk? Greenfield infrastructure in the Horn of Africa carries political, security and currency risk that commercial lenders price highly. Development finance institutions are the obvious candidates, and Afreximbank’s indicated willingness to support the wider network is relevant here.
- How is the tariff set? This is the question that determines whether Ethiopia gets cheaper fuel or an expensive fixed asset. A tariff designed to recover $660 million over fifteen years is a very different number from one designed to match the current trucking cost.
None of these has been answered publicly. Until they are, the project should be read as an announcement rather than a transaction.
What to watch
| Indicator | Why it matters |
|---|---|
| A published cost breakdown | Until port, storage, border and pipe are costed separately, the $5.5m/km figure cannot be benchmarked and the project cannot be assessed on cost. |
| The tariff framework | Determines whether the pipeline delivers cheaper fuel to Ethiopia or a return on capital to its owners. Those are different outcomes. |
| The ownership vehicle and the Djibouti government’s role | A pipeline originating in Djibouti requires Djibouti’s consent and participation. The structure will indicate how the two governments have divided the benefit. |
| Whether the inland run is also addressed | The pipeline stops at the border. The Ethiopian inland leg remains a trucking operation. Any plan to extend to Addis would change the economics substantially. |
| Ethiopia’s parallel diversification progress | If Berbera or Port Sudan become viable during construction, the strategic case for the pipeline weakens. |
Frequently asked questions
What is the Ethiopia–Djibouti pipeline?
A 120-kilometre refined product pipeline planned to run from Damerjog in Djibouti to Dewele on the Ethiopian border, at a reported cost of $660 million. It is the smallest of the three assets in Dangote’s announced African distribution network and is designed to replace the trucking and transhipment segment between the port and the border.
Why does the cost per kilometre look high?
Because the $660 million figure almost certainly covers more than pipe. A functional pipeline at Damerjog requires port and marine works, storage at the origin, border and customs infrastructure at Dewele and loading facilities on the Ethiopian side. No component breakdown has been published, so the per-kilometre figure cannot be benchmarked against a pure pipeline cost.
Why is Ethiopia dependent on Djibouti?
Ethiopia is landlocked and has no refinery of meaningful scale. Essentially all of its refined product imports arrive by ship at a foreign port and travel inland by road or rail. The Djibouti corridor has historically carried the large majority of that trade — by most estimates well above 90% — which creates a single-route dependence for a country of more than 130 million people.
Does the pipeline reduce Ethiopia’s dependence on Djibouti?
No — it increases efficiency on the Djibouti route rather than reducing dependence on it. The pipeline still originates in Djibouti. Ethiopia has been pursuing alternative corridors including Berbera, Port Sudan and Lamu, and a more efficient Djibouti route makes those alternatives relatively less attractive in the short term.
What does the pipeline replace?
The port-to-border truck leg and the transhipment at Dewele. Product currently arrives at the port, is loaded onto trucks, driven to the border, cleared through customs, transhipped onto Ethiopian-registered trucks and then driven inland. The pipeline removes the first truck leg and one handover. The inland run from Dewele onward remains a trucking operation.
Has the project been financed?
No financing structure has been publicly disclosed. The project has been announced at a cost of $660 million. A cross-border pipeline requires an ownership vehicle acceptable to both Djibouti and Ethiopia, a construction risk arrangement and a tariff framework. None of these has been published.
How does this fit with the Walvis Bay and Kenya Pipeline assets?
All three are distribution assets in the same network, but they do different things. Walvis Bay adds storage capacity on the Atlantic coast in a market that already has alternatives. The Kenya Pipeline stake buys access to a route that already works. The Ethiopia pipeline is the only one that changes the physical structure of supply for a market with no alternative route.
Related reading on Dedon Tech: Dangote’s Kenya Pipeline stake: buying the tollgate, not building a road · Dangote in Namibia: what the $2.5bn tank farm actually commits · right of way: why Nigerian fibre costs more than it should · full site map
0 Comments: