The Sovereign Aqueduct Network

Turning wasted monsoon and inland flood water into sovereign national wealth — 30,000 GL per year captured from water Australia currently loses to the ocean, stored at Alice Hub, and delivered by gravity to the south. No large northern dams. Revenue from AI campus water supply pays for the build.

1. Why Northern Water Has Been Rejected Before

The idea of redirecting northern Australia’s monsoon runoff to the dry south is not new. It has been proposed in various forms since John Bradfield outlined his scheme in the 1930s and 1940s. It has been studied, modelled, and rejected multiple times since. The most recent and comprehensive rejections come from three authoritative sources:

• CSIRO Bradfield Scheme Assessment (2021–22): Concluded the scheme was not economically viable and carried significant environmental risk.
• National Water Reform Advisory Council (NAWRA, 2016–19): Found that large-scale northern water transfer was not cost-competitive with alternatives and raised serious environmental concerns.
• Garnaut Panel: Identified evaporation losses, cost of infrastructure, and ecological disruption as fatal constraints.

These reports are credible, rigorous, and correct — about the scheme they assessed. The scheme they assessed bears almost no resemblance to the Sovereign Aqueduct Network. Every one of the fatal objections raised in those reports was a consequence of specific design assumptions — large dams in the north, continuous extraction from rivers, storage in tropical environments with high evaporation, and revenue modelled only from low-value irrigation. Change those assumptions and the objections dissolve.

The detailed comparison between what was assessed and what is proposed here is in Section 7. The rest of this memo describes what the Sovereign Aqueduct Network actually is.

2. The Scale of What Is Being Wasted

Northern Australia is one of the wettest inhabited regions on Earth — for four months of the year. The wet season delivers rainfall volumes that are difficult to comprehend at human scale.

The 30,000 GL target is a deliberately conservative fraction of what is available. Australia’s total urban and rural water use is approximately 70,000–80,000 GL per year. The network targets less than half that volume from a source that currently flows entirely to the ocean. The constraint on the target is not the water available — it is the infrastructure capacity to move and store it, which scales with the MMC corridor build programme.

The Lake Eyre Basin dimension. The inland contribution is often overlooked in northern water discussions because it is episodic rather than annual. But the Lake Eyre mega-floods — driven by rainfall across the Todd, Finke, Georgina, Diamantina, and Cooper Creek systems — deliver billions of GL that currently spread across inland floodplains and evaporate over weeks and months. The 2025–26 flood event alone delivered volumes that dwarf anything the proposed capture network targets. During these events, threshold pumps on inland channels capture overflow directly into the nearest MMC corridor pipeline, adding a drought-resilient supplement to the coastal capture system. This dual-source architecture — coastal rivers plus inland floods — provides far greater yield resilience than any single-source northern water scheme.

3. The Capture System — Two Coastal Types Plus Inland

The Sovereign Aqueduct Network uses three capture methods, selected to match the hydrology of each site. All three share the same operational rule: pumping occurs only during high-flow and flood events, powered by solar via the corridor HVDC lines, with strict environmental flow thresholds enforced at every site by real-time AI monitoring.

3.1 Type 1 — Overflow Ring Dam Capture

Small, low-impact overflow structures or ring tanks are positioned on floodplains adjacent to rivers. These structures fill only during major flood events when the river exceeds its banks onto the floodplain. Once the ring dam is full, solar-powered pumps run continuously until the tank is emptied into the feeder viaduct pipeline. The ring dam then sits empty, ready for the next event.

These are not large dams. They are off-channel storage ponds — no dam wall across the river, no barrier to fish passage, no disruption to the river’s natural flow regime. They are filled by floodplain overflow — water that would otherwise spread across country and evaporate. The environmental footprint of each structure is local and limited.

3.2 Type 2 — Direct River Threshold Pumping

No dam at all. Automated pump stations are positioned on the riverbank with a set activation threshold — a river flow level above which pumping begins and below which it stops. When the river runs in high flood, the pumps run. When the flow drops below the threshold — which is set to protect environmental base flows and ecological water requirements — the pumps stop automatically.

This method leaves the river physically unmodified. There is no structure in the river channel. The pump stations sit on the bank. Dry-season flows, low-flow ecological events, and the full natural hydrograph below the threshold are untouched. Only the peak flood excess — water that would continue to the ocean regardless — is captured.

3.3 Lake Eyre Basin Inland Capture

During episodic mega-floods in the Lake Eyre Basin, rivers including the Todd, Finke, Hugh, Georgina, Diamantina, and Cooper Creek produce massive overflow volumes that currently spread across inland floodplains and evaporate. Simple overflow weirs or threshold pump stations on key inland channels during flood events capture this overflow and pump it directly into the nearest MMC corridor pipeline or short spur connections to Alice Hub.

No new major northern infrastructure is required for inland capture. The MMC corridors themselves pass close to or through the Lake Eyre Basin catchment. The inland capture system supplements coastal capture during flood years and provides additional drought-year resilience by capturing events that are geographically closer to Alice Hub — requiring shorter feeder distances and lower pumping energy.

3.4 Lake Argyle — the headline node already built

Lake Argyle on the Ord River in the East Kimberley is the headline finger of SBC Corridor 5 (the Kimberley corridor). It is Western Australia's largest and Australia's second-largest man-made freshwater reservoir — 10,763 GL at full supply, with a maximum flood capacity of 35,000 GL following the 1996 spillway raise. The catchment is 46,100 km². At full supply, the lake holds 21 times the volume of Sydney Harbour. The infrastructure is already built. The reservoir, the dam wall, the spillway, the regulating valves, and the flood-storage capacity were all completed by 1972, with the 1996 expansion adding the additional flood headroom.

The Ord River Irrigation Area Stage 1 is allocated 335 GL per year — the entire current irrigation entitlement against a reservoir of more than 10,000 GL. The Western Australian government's 2024 Ord River Irrigation Area Strategy plans expansion to 50,000 hectares (up from approximately 28,000 hectares today) over the 10-year horizon. Even with the planned expansion, the irrigation allocation remains a fraction of the reservoir volume and an even smaller fraction of annual inflow in wet years. In recent wet seasons, the dam has experienced multi-month spillway overflows discharging directly into the Cambridge Gulf. The January 2025 overflow was the first on record at that date in the season, following heavy early monsoon rains that filled the reservoir to over 100% capacity. The water discharges to the ocean.

The Ord region has also proven a structurally difficult agricultural environment despite over fifty years of intensive investment. Early cotton crops in the 1960s and 1970s failed under tropical pest pressure. The Quintis sandalwood operation, the largest single agricultural investor in the region for the past two decades, entered receivership in April 2024. Distance from markets, tropical pest dynamics, climate volatility, and tropical soil chemistry have all made consistent agricultural profit difficult to sustain. The water at Lake Argyle produces higher economic value when moved south through the corridor network and applied in established southern agricultural zones — the Murray-Darling Basin under drought-proofed conditions, agrivoltaic on the corridor, and the central Australian Solar Region — than it does in expanded local irrigation at the dam itself.

For the Sovereign Aqueduct Network, Lake Argyle is one finger but a major one. The dam is already built. The surplus is already verified by the spillway overflow record. The connection — the elevated viaduct feeder running south from Lake Argyle to the Phase 3 Kimberley Corridor and onward to Alice Hub — is what the MMA programme adds. The argument that “there is not enough water in the north” has to contend with one existing dam discharging multi-month spillway overflows to the Cambridge Gulf in recent wet years. Lake Argyle does not replace the need for direct river threshold pumping from the Fitzroy, Daly, Victoria, Roper, Mitchell, Flinders, Gilbert, and other northern rivers — the network needs all the fingers. But Lake Argyle is the single most powerful single-node counter to the “no water in the north” objection, because the dam is already built and the surplus is already running to the ocean.

3.5 Operational Rules

All three capture types operate under the same strict framework:

• Wet season and flood events only. No pumping during the dry season. No continuous extraction. The system is idle for six to eight months of the year.
• Environmental threshold enforcement. Every pump station has a set minimum river flow below which pumping is prohibited. These thresholds are set by hydrological assessment of ecological flow requirements at each site and enforced automatically by the monitoring system.
• Solar-powered pumping. All pump energy comes from the corridor HVDC solar supply. Zero diesel. Zero grid draw. Operating cost is near-zero once the infrastructure is built.
• Real-time AI monitoring. Flow rates, pump status, environmental thresholds, and water quality are monitored continuously at every capture site. Anomalies trigger automatic pump shutdown. The entire network is managed from a central operations facility.
• Addition, not reallocation. The network captures water that would otherwise reach the ocean. It does not reduce flows available to downstream users, ecosystems, or the coastal environment below the threshold levels. This is new water entering the productive economy — not water taken from existing users.

4. The Elevated Viaduct Delivery Network

The most distinctive engineering feature of the Sovereign Aqueduct Network is its delivery system: an elevated viaduct network rather than buried pipelines or ground-level channels. This choice is not aesthetic. It follows directly from the engineering requirements of the MMC corridor system and resolves the environmental, construction, and maintenance objections that have plagued previous northern water proposals.

4.1 The Standard Feeder Viaduct

Each feeder viaduct branches from a capture site to the nearest MMC main corridor, carrying multiple services on a single structure:

• Main water pipeline — 1.8–2.0 m diameter, sized for peak intermittent flow during flood events
• HVDC and power transmission lines — powering the pump stations at the river end
• Fibre optic and control cables — real-time monitoring and automated pump management
• Service rail — narrow-gauge or industrial rail running the full length of every feeder for construction logistics and permanent maintenance access
• Maintenance walkway — for inspection and manual intervention where required

Once water reaches the main MMC corridor via the feeder viaduct, it travels along the corridor’s own pipeline to Alice Hub. The main corridor pipeline is a shared-infrastructure asset — it carries water from every feeder viaduct along its route, aggregating the capture from dozens of river systems into a single delivery flow.

4.2 The Self-Building Construction Method

The feeder viaducts are built using the same incremental “install-as-you-go” method as the main MMC corridors. All foundation drilling, pier erection, deck installation, and service fit-out is performed from the already-completed section of viaduct ahead of the construction front. The service rail on the completed deck carries the construction equipment and materials forward.

This eliminates the need for temporary ground roads through floodplains, wetlands, and remote country. There is no network of access tracks bulldozed through intact ecosystems. The construction footprint is the viaduct itself — and that footprint is elevated above the ground. When the viaduct is complete, the country underneath it is unchanged. The construction phase leaves behind the viaduct and nothing else.

Early sections of the viaduct become operational while later sections are still under construction. The first capture sites along a feeder route begin delivering water to Alice Hub before the feeder reaches its furthest river. The system generates revenue while it builds.

4.3 Why Elevated Is the Right Choice

5. Central Storage at Alice Hub — The Design Principle That Changes Everything

The single most important design decision in the Sovereign Aqueduct Network is where the water is stored. The answer is not in the north. It is at Alice Hub.

Every past northern water scheme has proposed storing water close to where it falls — in large dams on northern rivers, in tropical reservoirs, or in floodplain impoundments. This creates three fatal problems that the official reports correctly identified: evaporation losses in tropical environments are enormous (up to 2,000 mm per year from open water surfaces), large northern dams carry extreme ecological risk to river systems and dependent ecosystems, and the cost of building and maintaining large dams in remote tropical environments is prohibitive.

The Sovereign Aqueduct Network eliminates all three problems by moving water quickly out of the tropics and storing it at Alice Hub.

5.1 Why Alice Hub

Alice Hub sits at approximately 550 m elevation in the MacDonnell Ranges — the geographic and hydrological centre of the continent. Storage at this elevation provides several decisive advantages:

• Lower evaporation. The arid inland environment at Alice Hub has significantly lower annual evaporation than tropical storage sites. Covered or PHES-integrated storage reduces surface evaporation further still. Evaporation losses that would consume 20–30% of tropical stored water are a fraction of that at Alice.
• Gravity distribution south. Water stored at 550 m elevation gravity-feeds south toward South Australia, the Murray-Darling Basin, and the populated southeast — at near-zero ongoing energy cost. The pumping energy is expended once, moving water from the river to the corridor pipeline. After that, physics does the work.
• PHES integration. The Alice Hub pumped hydro energy storage system uses water as its working fluid. The Sovereign Aqueduct Network replenishes PHES reservoir volumes, extending the system’s energy storage capacity as the water volume grows. Water and energy storage are integrated into the same physical infrastructure.
• Multi-purpose demand hub. Alice Hub is the convergence point for AI campus water demand, solar farm cooling, corridor town supply, hydrogen production, and southern agricultural gravity delivery. All demand streams are served from a single, well-managed central reservoir rather than dispersed across dozens of remote northern sites.
• No large northern dams. Because all storage is at Alice Hub, no large dam structures are needed in the north or on remote inland floodplains. The capture structures — ring tanks and pump stations — are small. The feeder viaducts carry water out of the north quickly. The north is left structurally intact.

5.2 The Phased Corridor Build

The network builds in phases aligned with the MMC corridor construction programme:

• Phase 1 — Darwin–Alice Corridor: Priority start. Captures from 10–20 Top End rivers and early Lake Eyre Basin events. Immediate revenue from Alice AI campuses and solar farms. Gravity delivery south to South Australian users begins as soon as Alice Hub storage reaches operational levels.
• Phase 2 — Gulf Corridor: Additional capture from the Mitchell, Flinders, Gilbert, and other Gulf of Carpentaria rivers. Significantly increases annual yield from the Gulf’s consistently high-flow systems.
• Phase 3 — Kimberley Corridor: Lake Argyle as the headline node (see §3.4) plus direct threshold pumping from the Fitzroy, Daly, Victoria, and other Kimberley rivers. Western Australia’s northern rivers carry some of the highest annual flow volumes in the country, providing a large additional capture base on top of the existing Lake Argyle surplus.

6. The Revenue Cascade — Water Sold at Commercial Rates

The Sovereign Aqueduct Network is not an irrigation scheme funded by government to support agriculture. It is a water utility that sells water at commercial rates to a priority cascade of high-value users, using that revenue to service the infrastructure construction debt.

The priority order is determined by value per kilolitre delivered:

6.1 First Priority — AI Data Centre Cooling

AI data centres are the highest-value water consumers in the cascade. A 1 GW AI campus using evaporative cooling consumes 1–3 billion litres per year. At commercial industrial water rates, ten such campuses along the MMC corridor generate hundreds of millions of dollars per year in water revenue alone — before power and fibre revenue are counted. This revenue stream begins as soon as the first Alice Hub AI campus is operational. It is the financial engine that makes the aqueduct network self-funding.

The water demand from AI campuses is also the most predictable and contractable. A hyperscaler signing a 20-year take-or-pay water supply agreement provides the bankable revenue that secures the construction finance for the feeder viaduct network. The AI campus is not just the first customer — it is the anchor customer whose contract makes every downstream water user’s supply possible.

→ See: Australia’s AI Power Play — MMA Memo 8. The Power Imperative — MMA Memo 13.

6.2 Second Priority — Solar, Hydrogen, and Industrial

Large-scale solar PV requires modest water volumes for panel cleaning and substation cooling. Green hydrogen production via electrolysis requires significant water volumes — approximately 9 litres of high-purity water per kilogram of hydrogen produced. At the scale of the Alice Hub solar and hydrogen programme, this represents a substantial and contracted water demand. Industrial process water for the corridor’s manufacturing hubs is similarly predictable and commercially priced.

6.3 Third Priority — Corridor Towns and Domestic Supply

The 200 corridor towns and 11 intersection cities along the MMC network receive water supply from the aqueduct as a design feature of the corridor — not as a separate water infrastructure programme. Town water supply is a high-reliability, politically important use that is given third-priority allocation after the contracted commercial users.

6.4 Fourth Priority — Southern Irrigation

Agricultural irrigation — the primary focus of all past northern water proposals — receives fourth-priority allocation in the Sovereign Aqueduct Network. This is not a demotion of agriculture’s importance. It is a financial architecture that ensures the high-value commercial users fund the infrastructure that then delivers drought security to the agricultural sector at a cost they can afford.

Southern irrigation users receive gravity-fed water from Alice Hub at a fraction of the cost of alternative water supply sources. The AI campus revenue pays for the aqueduct. The farmers benefit from an infrastructure they could not have funded themselves. This is the economic logic that makes the Sovereign Aqueduct viable where the Bradfield Scheme was not: the high-value anchor tenants at Alice Hub pay for the infrastructure that the agricultural sector has always needed but could never justify on irrigation economics alone.

7. Why This Is Not the Bradfield Scheme

The comparison to the Bradfield Scheme is inevitable and worth addressing directly. The Bradfield Scheme proposed diverting northern Queensland rivers inland and south to irrigate the dry interior — a vision of transforming Australia’s climate through engineering. Every assessment since has found it unviable on environmental, economic, and hydrological grounds. Those assessments are correct — about the Bradfield Scheme. The Sovereign Aqueduct Network is a different proposal in every material respect.

The CSIRO, NAWRA, and Garnaut assessments correctly rejected a scheme that proposed large tropical dams, continuous extraction, and irrigation-only revenue. They did not assess — and their conclusions do not apply to — a system that proposes no large northern dams, flood-only threshold capture, solar-powered intermittent pumping, elevated viaduct delivery, central arid storage, and AI campus revenue as the primary financial anchor. The Sovereign Aqueduct Network is not a revised version of the Bradfield Scheme. It shares only the geographic context.

Lake Argyle (see §3.4) reinforces the point. The Bradfield Scheme required building large new northern dams — the central reason it was rejected. The MMA programme requires none. Lake Argyle is already built, has been operational since 1972, has had its flood storage capacity expanded once already (the 1996 spillway raise), and is currently discharging multi-month spillway overflows to the Cambridge Gulf in wet years. The MMA programme adds the connection — the elevated viaduct from Lake Argyle south to the Phase 3 Kimberley Corridor — not the dam. Where direct river threshold pumping is used from the Fitzroy, Daly, Victoria, and other northern rivers, no dam is built at all. The Bradfield objection to large northern dams is correct, and the MMA programme respects it by design.

8. Environmental and Indigenous Safeguards

The Sovereign Aqueduct Network captures water that Australia currently loses. It does not reduce flows to any existing user, ecosystem, or coastal environment — because the threshold capture rules ensure that only peak flood excess is taken. Below the environmental threshold at every site, the river runs as it always has. The safeguards are not an add-on to the programme. They are structural features of its design.

8.1 Environmental Flow Protection

Every capture site has a site-specific environmental flow threshold determined by hydrological assessment of the river’s ecological requirements — the flows required to maintain fish passage, riparian vegetation, wetland function, estuarine dynamics, and the cultural values of water for Traditional Owners. These thresholds are set conservatively and enforced automatically. Pumping stops the moment flow drops to the threshold. The system has no manual override for commercial reasons — environmental compliance is hardwired.

Real-time AI monitoring of flow rates, water quality, and ecological indicators at every site enables adaptive management. If monitoring reveals unexpected ecological response at any site, that site’s threshold is raised automatically and reviewed by the environmental management team. The system learns from its own operation.

8.2 Indigenous Co-Design and Benefit-Sharing

The feeder viaduct network crosses country that is the country of First Nations peoples who have managed these river systems for tens of thousands of years. Their knowledge of river behaviour, flood patterns, ecological relationships, and culturally significant sites is the most detailed and long-standing knowledge available — and it is indispensable to the design and operation of the capture system.

The Sovereign Aqueduct Network commits to full Traditional Owner co-design at every stage: site selection, environmental threshold setting, construction planning, and operational management. Traditional Owner benefit-sharing — through equity in the Sovereign Corridor Trust, employment in construction and operations, and preferential water allocations for community and cultural uses — is a design requirement, not a consultation afterthought.

The elevated viaduct design supports this commitment. Its small footprint, its ability to route around areas of significance, and its permanent maintenance access from the deck rather than from ground-level roads all reduce the long-term presence on country compared to any alternative delivery method.

8.3 This Is Addition, Not Reallocation

Every past northern water proposal has faced opposition from downstream users, coastal communities, and environmental advocates on the grounds that taking water from northern rivers reduces what is available to the ecosystems and communities that depend on them. That concern is legitimate — for schemes that propose continuous extraction or dry-season diversion.

The Sovereign Aqueduct Network captures flood overflow that currently reaches the ocean. It does not reduce what is available to downstream users, coastal communities, or estuarine ecosystems because the threshold rules ensure those flows are protected first. The network adds water to the productive economy that is currently lost entirely. It is genuinely additional supply — and that distinction is the foundation of its environmental and political sustainability.