Rivers of wasted power: How New Zealand's three great hydro systems underperform their physical potential — and who benefits from the gap
- Grant McLachlan
- 5 hours ago
- 41 min read
New Zealand's Waikato, Waitaki, and Clutha river systems account for 38% of national electricity generation. A physics-based audit of all three reveals a consistent pattern: obsolete turbines running decades without replacement, structural market incentives that reward withholding water over dispatching electricity, and publicly built infrastructure managed under rules that serve shareholder returns ahead of renewable output. The combined recoverable gap — generation already latent within existing dams and water rights — is approximately 3,100 GWh per year. That is enough to eliminate coal generation in a drought year and to permanently reduce the electricity price spikes that cost New Zealand consumers hundreds of millions of dollars annually.
Contents
1. The Waikato River and Tongariro Power Scheme
1.1. The spine of North Island power
1.2. The physics of a falling river
1.3. The staircase: hydraulic cascade and generation data
1.4. The Tongariro Power Scheme: the hidden tributary
1.5. Are the flows producing maximum electricity?
1.6. Wasting water to raise prices: the market gaming problem
1.7. Obsolete turbines: Mercury's own evidence
1.8. What a reformed Waikato system could produce
Figure 1 — Hydraulic elevation profile: Tongariro & Waikato
2.1. Introduction: the South Island's largest power system
2.2. The physics of the Waitaki cascade
Table 2A — Waitaki hydraulic staircase
Figure 2 — Waitaki hydraulic elevation profile
2.3. The 2011 Tekapo split: replicating the Genesis Pincer
2.4. Benmore's efficiency gap and the cost of deferred upgrades
2.5. Project Aqua and the road not taken
2.6. Waitaki: national share and quantified potential
3.1. Introduction: New Zealand's largest river by volume
3.2. The physics of the Clutha
Table 3A — Clutha hydraulic staircase
Figure 3 — Clutha / Mata-Au hydraulic elevation profile
3.3. The Roxburgh scandal: 68 years without a turbine upgrade
3.4. The Contact Energy conflict: a third pincer
3.5. Sediment: the existential threat
3.6. Clutha: national share and quantified potential
4.1. Three rivers: a national electricity audit
4.2. Conclusion: the privatisation dividend that never was
Executive Summary
This report analyses the physical performance, market dispatch behaviour, and structural ownership conflicts of New Zealand's three largest hydroelectric systems: the Waikato River cascade and Tongariro Power Scheme (operated by Mercury Energy and Genesis Energy respectively); the Waitaki Hydro Scheme in the Mackenzie Basin (Meridian Energy and Genesis Energy); and the Clutha / Mata-Au scheme in Central Otago (Contact Energy). Together, these systems produce approximately 16,790 GWh annually — 38.3% of New Zealand's national electricity generation.
The central finding is that each of these systems produces substantially less electricity than the physics of the water flowing through them permits — and that this shortfall is not primarily a consequence of natural constraints or engineering limitations, but of a market structure that systematically rewards the management of scarcity over the maximisation of renewable generation.
Key Finding 1 — Turbine efficiency losses: ~$1.8 Billion in permanently foregone generation
Across all three systems, turbines were operated for decades — in some cases more than 65 years — without replacement or significant refurbishment. The evidence that this deferred investment caused measurable, quantifiable efficiency losses comes from the operators themselves: every upgrade announcement has documented the GWh gain attributable to improved efficiency, confirming that the pre-upgrade plant was underperforming. Mercury's Karāpiro refurbishment delivered +32 GWh from a 5% efficiency gain. Meridian's Benmore refurbishment delivered +70 GWh from a 6% efficiency gain. Contact Energy's partial Roxburgh upgrade is projected to deliver +44 GWh from four of eight turbines. Each of these figures allows a precise calculation of the generation permanently lost during the deferral period.
Table ES-1 — Permanently Foregone Generation from Deferred Turbine Investment
System | Station | Commissioned | Upgrade (or first upgrade) | Deferral (years) | Efficiency deficit | GWh permanently foregone (est.) | Value @ $120/MWh |
Waikato | Karāpiro | 1947 | 2019–22 | ~75 yrs | 5% | ~1,893 GWh | ~$227M |
Waikato | Ātiamuri | 1956 | 2031–34 (sched.) | ~75 yrs at upgrade | 10.5% | ~2,412 GWh | ~$289M |
Waikato | Maraetai I | 1953 | 2030–34 (sched.) | ~77 yrs at upgrade | ~4.8% | ~3,493 GWh | ~$419M |
Waitaki | Benmore | 1965 | 2008–10 | 43 yrs | 6% | ~5,660 GWh | ~$679M |
Waitaki | Ōhau B | 1984 | No upgrade announced | 40+ yrs (ongoing) | ~3% (est.) | ~1,165 GWh to date | ~$140M |
Waitaki | Ōhau C | 1985 | No upgrade announced | 40+ yrs (ongoing) | ~3% (est.) | ~1,165 GWh to date | ~$140M |
Clutha | Roxburgh | 1956–61 | 2024–26 (4 of 8 turbines) | 68 yrs | ~3% (est.) | ~3,290 GWh | ~$395M |
TOTAL |
|
|
|
|
| ~19,078 GWh | ~$2.29B |
Table ES-1 — Permanently foregone generation from deferred turbine investment across the three systems. 'Permanently foregone' means generation lost during the deferral period that cannot be recovered. Efficiency deficit percentages are derived from operators' own upgrade announcements. Ongoing losses (Ōhau B/C, Roxburgh remaining 4 turbines) continue to accumulate. Clutha Roxburgh 4 remaining turbines not yet scheduled. Value is at $120/MWh wholesale. Sources: Mercury Energy, Meridian Energy, Contact Energy, International Water Power.
Key Finding 2 — Untapped head: Generation flowing to the sea
Each of the three systems has structural gaps — reaches of river where significant head exists but no turbines have been installed — that represent latent renewable generation capacity far exceeding any wind or solar project currently under construction in New Zealand. These gaps are not all recoverable: some are environmentally protected, others require investment on a scale that the market has repeatedly failed to justify. But their combined scale illustrates the magnitude of the country's untapped renewable endowment within already-consented, already-used river systems.
Table ES-2 — Untapped Head: Structural Generation Gaps
System | Gap / location | Head (m) | Potential capacity / output | Status |
Waikato | Huka Falls (Taupō outlet) | 11 m | ~33 MW continuous | Environmentally protected. Tourism site. Not developable. |
Waikato | Whakamaru→Maraetai inter-station gap | 35 m | ~114 MW | Geographically uncaptured since 1920s. No dam site between stations. ~114 MW × 8,760 hrs ≈ ~999 GWh/yr theoretical. |
Waikato | Other inter-station free-fall gaps | ~15 m total | varied | Multiple smaller gaps in upper cascade. Structurally uncaptured. |
Waitaki | Lower Waitaki (Project Aqua site) | ~200 m | ~520 MW / ~4,500 GWh/yr | $38.7M spent on consenting before cancellation in 2004. North Bank Tunnel: water rights 2009, suspended 2013. Still undeveloped. |
Clutha | Lake Hāwea control dam (no turbine) | 70 m | ~250 MW / ~2,200 GWh/yr | Dam built 1958 as pure storage control structure. No power generation ever installed. 70 m of head permanently wasted. Contact Fast-Track proposal raises Hāwea +2 m but still no turbine proposed. |
Clutha | Upper Clutha free river (Wānaka→Clyde) | 46 m | ~180 MW | Entirely undeveloped. Kawarau R. (from Wakatipu) joins at Cromwell. Long-recognised as viable dam site. No proposals currently active. |
Clutha | Lower Clutha (Roxburgh→sea) | ~112 m | ~2,700 GWh/yr theoretical | 160 km to Balclutha. Multiple former dam proposals, all abandoned. Ecologically significant braided river habitat. Unlikely to be developable under current RMA and environmental policy. |
Table ES-2 — Structural generation gaps across the three systems. 'Head' is the elevation drop available at each gap. Potential output figures are theoretical maximums at modern turbine efficiency. Not all are developable; environmental, cultural, and ecological factors are legitimate constraints. The Hāwea figure is particularly striking: 70 m of head across a dam that already exists, built in 1958, where no power station has ever been installed. Sources: Klaut Media analysis; Meridian Energy; Contact Energy; Wikipedia; MBIE.
Key Finding 3 — Market inefficiencies: The structural incentive to withhold
Each of the three systems exhibits a version of the same market distortion: a structure in which the commercial incentive for at least one operator is to withhold renewable generation rather than maximise it. The mechanisms differ in form — the Genesis pincer, Meridian's undesirable trading practices, Contact's thermal portfolio — but they share a common architecture: an operator that controls both a renewable generation asset and a complementary position that benefits when the renewable asset is withheld. The market was designed to prevent any single company dominating generation. It succeeded in preventing market domination while creating market manipulation.
Table ES-3 — Market Inefficiencies and Structural Conflicts of Interest
System | Mechanism | How it works | Quantified impact |
Waikato | Genesis Pincer — Tongariro/Huntly | Genesis controls Tongariro inflows into Lake Taupō (upstream tap) AND the Huntly thermal station (downstream price floor). Modulating Tongariro diversions downward causes Mercury to conserve Taupō storage, triggering Huntly thermal dispatch at elevated spot price. Genesis receives clearing price on both Tongariro generation and Huntly thermal. | August 2024 spot price reached $800/MWh — 7–10× typical level. MBIE independent review concluded market 'lacks reliable back-up options during dry years.' Annual consumer excess cost in dry years estimated at $780M–$1.56B. |
Waikato | Maraetai Dispatch Withholding | Maraetai I & II operate at 27% capacity factor — anomalously low for a station with a full reservoir and no ecological minimum flow constraint. Dispatching Maraetai freely would add ~680 GWh/yr and reduce dry-period price spikes. | ~680 GWh/yr withheld ≈ $82M/yr at $120/MWh. At peak $800/MWh event pricing, the same withheld generation would be worth ~$544M/yr to consumers foregone. |
Waitaki | Meridian Undesirable Trading | In a landmark ruling, the Electricity Authority found Meridian engaged in undesirable trading practices by spilling water from Waitaki lakes rather than generating electricity. Water was physically disposed of because the spot market made spilling more profitable than generating, given Meridian's retail hedge book position. | Electricity Authority ruling: consumer market costs in excess of $80M from a single episode of strategic spilling. Ruling also confirmed the behaviour was a rational market response — meaning it is repeatable. |
Waitaki | Genesis Tekapo Split — South Island Pincer | 2011 Government transfer of Tekapo A&B to Genesis replicated the Tongariro/Waikato architecture. Genesis now controls the Tekapo upstream tap into Lake Pūkaki — New Zealand's national electricity battery — while Meridian manages downstream storage and dispatch. Neither company is accountable for total system output. | Lake Pūkaki holds 1,767 GWh stored energy — equivalent to 9,000 grid-scale batteries. It is the single largest leverage point over national spot prices. Unquantified but structural; any Pūkaki withholding sets marginal price for entire national HVDC grid. |
Clutha | Contact Energy Thermal Conflict | Contact owns both Clyde + Roxburgh (combined ~3,700 GWh hydro) and the Taranaki Combined Cycle gas station (~550 MW). Contact can modulate Lake Hāwea releases within resource consent conditions, reducing hydro output and triggering Taranaki CC dispatch at elevated spot price. No collusion required — the market structure produces the outcome automatically. | Fast-Track application to raise Hāwea by +2 m increases strategic storage leverage simultaneously. No regulatory mechanism currently prevents Contact managing its renewable and thermal assets as an integrated profit-maximising portfolio. |
Table ES-3 — Market inefficiencies and structural conflicts of interest across all three systems. Each involves an operator that holds both a renewable asset and a complementary position benefiting from renewable withholding. The behaviour in each case is commercially rational within the rules of the half-hourly spot market. The rules, not the operators, are the source of the problem. Sources: Electricity Authority; MBIE; Klaut Media analysis.
Key Finding 4 — The combined recoverable gap
Taking conservative estimates — excluding all structural head gaps requiring new infrastructure, including only efficiency upgrades comparable to those already proven at Karāpiro and Benmore, and assuming optimised dispatch without market withholding — approximately 3,100 GWh per year of additional renewable generation is recoverable from these three systems without building a single new dam, securing new water rights, or significantly expanding the environmental footprint of any scheme. This is not a theoretical maximum. It is a conservative estimate of what better turbines and rational dispatch rules would produce from water already flowing through dams already built.
~3,100 GWh per year — recoverable without new dams, new water rights, or new consents.
At 43,879 GWh national consumption (2024), this 3,100 GWh represents 7% of national generation: enough to power approximately 370,000 additional average New Zealand homes, to eliminate coal-fired electricity in a normal drought year, and to permanently reduce the risk of the $800/MWh price spikes that dominate annual average consumer costs. The analysis that follows provides the physics-based evidence for each of these claims, system by system.
1. The Waikato River and Tongariro Power Scheme
1.1. The spine of North Island power
The Waikato River is, in engineering terms, a perfect machine. It falls roughly 302 metres from the lip of Lake Taupō — itself sitting at 357 metres above sea level — to the tailrace of Karāpiro Dam at approximately 55 metres above sea level, before flowing on to the Tasman Sea at Port Waikato. That 302-metre descent, distributed across eight dams and nine generating stations over a river distance of 188 kilometres, underpins approximately 13% of New Zealand's total annual electricity generation, or roughly 4,000 gigawatt-hours (GWh) per year. Add the Tongariro Power Scheme, which diverts snowmelt and rainfall from the flanks of Mounts Ruapehu and Tongariro northward into Lake Taupō, and the combined system is the single most important electricity production asset in the North Island.
All nine Waikato stations are operated by Mercury Energy — listed on the NZX, majority-owned by the Crown. The two Tongariro stations (Rangipo and Tokaanu) are operated by Genesis Energy — also majority Crown-owned. The dams themselves were built between 1929 and 1971 as part of a government programme. They were operated as a unified system by the Electricity Corporation of New Zealand (ECNZ) until that corporation was broken up into three competing state-owned enterprises on 1 April 1999.
That 1999 split is the pivotal event in the story of the scheme's underperformance. It did not merely create competition between generators. It created a structural conflict of interest — a situation where one company, Genesis, controls both the upstream water supply to Lake Taupō and the thermal generators that set the spot price when hydro is withheld. This article analyses how that conflict works, quantifies the physical underperformance, and asks whether any regulator has the mandate or the tools to address it.
This article asks four questions the electricity market does not.
Given the volume of water flowing through the system, are the stations extracting the maximum power the physics permits?
How much of the river's potential elevation is actually being captured, and how much falls freely between turbines generating nothing?
Could modern turbines produce meaningfully more electricity from the same flow?
Most troublingly — is the market structure creating perverse incentives to withhold generation while one company profits from manufactured scarcity?
1.2. The physics of a falling river
The power output of any hydro station is determined by: P = ρ × g × Q × H × η, where P is power in watts, ρ is water density (1,000 kg/m³), g is gravitational acceleration (9.81 m/s²), Q is volumetric flow in cumecs, H is net head in metres, and η is combined turbine-generator efficiency (typically 0.87–0.93).
Simplified: P (MW) ≈ 0.00981 × Q (cumecs) × H (metres) × η
Modern Francis and Kaplan turbines achieve 90–93% efficiency at design conditions. Original mid-century turbines typically achieved 80–87%, declining further with age. This 6–13 percentage point gap represents, at Waikato scale, hundreds of gigawatt-hours of foregone annual generation from the same water. Mercury's own upgrade data confirms the gap is real and measurable: the Karāpiro turbine replacement delivered a 5% efficiency improvement and 32 GWh of additional annual output from identical flow conditions.
The Waikato River's mean discharge of 340 cumecs at Karāpiro sets the theoretical floor for generation calculations. The total gross head available — from Taupō at 357 metres to the Karāpiro tailrace at approximately 55 metres — is 302 metres. Only approximately 241 metres of that 302 metres is actively captured by turbines. The remaining ~61 metres falls through unimpounded river reaches between stations, generating no electricity.
1.3. The staircase: hydraulic cascade and generation data
The following two tables present a complete technical audit of the Waikato Hydro Scheme. Table 1A maps the entire cascade from Lake Taupō to Karāpiro as an elevation profile, showing for each station the upstream lake or headpond level, the dam crest elevation, the downstream tailrace level, and the resulting gap between the bottom of each dam and the crest of the next dam downstream.
That final column — the gap — is the article's central measurement. Where it is large, a section of river falls freely between two stations, converting potential energy into turbulence and heat rather than electricity. Where it is zero or negative, the two stations are hydraulically chained: the downstream headpond rises to meet the upstream tailrace, and every cubic metre of water passes from one turbine directly to the next with minimal waste.
All elevations are in metres above sea level (masl). Dam crest elevations are estimated as headpond full supply level (FSL) plus approximately 2 metres of freeboard. Design flow figures are back-calculated using P = ρgQHη at η = 0.87 for mid-century equipment.
Table 1A — Waikato/Tongariro Hydraulic Staircase: Lake Levels, Dam Crests, Tailrace Bases, and Inter-Station Gaps
Station / Feature | Headpond (masl) | Dam crest est. (masl) | Tailrace base (masl) | Net head (m) | Gap to next dam crest (m) | Capacity (MW) | Notes |
Lake Taupō (source) | 357 | — | 357 | — | 11 | — | 59B m³. Huka Falls = 11 m uncaptured |
Aratiatia | 346 | 348 | 319 | 27 | 0 | 90 | Run-of-river; 55 ha headpond |
Ōhākurī | 319 | 321 | 282 | 37 | 0 | 112 | Largest artificial lake on Waikato |
Whakamaru | 282 | 284 | 218 | 64 | 35 ▲ | 100 | Grid switching station; 31 m deep diversion channel |
Maraetai I & II | 183 | 185 | 142 | 40 | 2 | 360 | Arch dam. CF 27% — lowest in cascade. Two stations |
Waipapa | 140 | 142 | 115 | 25 | 0 | 54 | Lower cascade — hydraulically chained |
Arapuni | 115 | 117 | 85 | 30 | 0 | 197 | 64 m arch dam; largest Waikato dam |
Karāpiro | 85 | 87 | 55 | 30 | — | 112.5 | Upgraded 2019–22. +32 GWh, +5% efficiency |
Tongariro Scheme | various | — | →Taupō | ~300 | — | 360 | Genesis: Rangipo 120 MW + Tokaanu 240 MW |
TOTALS | 357→55 | — | — | 241 m captured | 61 m wasted | 1,075 MW | ~12.4% national |
Table 1A — Waikato/Tongariro Hydraulic Staircase. Red-highlighted cell: the 35-metre Whakamaru→Maraetai gap is the largest single piece of wasted head in the cascade — ~114 MW of continuously uncaptured potential. The lower four stations (Waipapa, Arapuni, Karāpiro, Maraetai) are hydraulically chained with near-zero inter-station loss. masl = metres above sea level. Sources: Mercury Energy; Engineering New Zealand; Wikipedia.
Table 1B — Waikato/Tongariro Generation Data: Capacity, Output, and Upgrade Status
Station | Commissioned | Capacity (MW) | Annual output (GWh) | Capacity factor | Operator | Key facts / upgrade status |
Aratiatia | 1964 | 90 | ~400 | 51% | Mercury | Run-of-river. No upgrade announced. 55 ha headpond. |
Ōhākurī | 1961 | 112 | ~400 | 41% | Mercury | Upgrade 2029–32: +10 GWh. Largest artificial lake on Waikato River. |
Whakamaru | 1956 | 100 | ~430 | 49% | Mercury | Grid switching station. 31 m deep concrete diversion channel. No upgrade scheduled. |
Maraetai I & II | 1953/1954 | 360 | ~945 | 27% ⚠ | Mercury | Arch dam. CF of 27% is anomalously low — commercial dispatch choice, not physical constraint. Upgrade 2030–34: +45 GWh. Oldest major turbines on scheme. |
Waipapa | 1961 | 54 | ~235 | 50% | Mercury | Lower cascade; hydraulically chained between Maraetai and Arapuni. |
Arapuni | 1929 | 197 | ~790 | 46% | Mercury | 64 m concrete arch dam — largest on Waikato. NZ heritage site. No upgrade scheduled. |
Karāpiro | 1947 | 112.5 | ~505 | 51% | Mercury | $90M upgrade completed 2022. +32 GWh, +5% efficiency. 3 Kaplan turbines. Min 148 cumecs ecological flow. De facto baseload. |
Rangipo (Tongariro) | 1983 | 120 | ~600 | 57% | Genesis | Underground cavern, 6.2 km tunnel. Transfers water from Tongariro catchment to Lake Taupō. |
Tokaanu (Tongariro) | 1973 | 240 | ~800 | 38% | Genesis | Largest Tongariro station. 4 × 60 MW. Transfers plateau water to Taupō. |
TOTAL |
| 1,075 MW | ~4,049 GWh | 43% avg | Mercury + Genesis | Committed upgrades: +119 GWh p.a. by 2034. Further identifiable potential: ~445 GWh p.a. |
Table 1B — Waikato/Tongariro generation data. ⚠ Maraetai I & II capacity factor of 27% is anomalously low and reflects market dispatch strategy rather than any physical or hydrological constraint. Sources: Mercury Energy; Wikipedia; Engineering New Zealand.
Several features of Table 1A demand immediate attention. The most striking is the Whakamaru–Maraetai gap: the river exits Whakamaru station's tailrace at approximately 218 masl and does not encounter another turbine until Maraetai's headpond at approximately 183 masl — a free-fall of 35 metres over several kilometres of unimpounded river. At 370 cumecs mean flow, this gap represents a continuous uncaptured power potential of approximately 9.81 × 370 × 35 × 0.90 ≈ 114 MW. It is not commercially or environmentally feasible to recover all of this. But its magnitude illustrates the structural incompleteness of a cascade that was never fully engineered to capture its own available head.
Third: the Huka Falls. The 11-metre drop between the Taupō outlet gates (357 masl) and the Aratiatia headpond (346 masl) is the most famous uncaptured head in the system. A run-of-river station at Huka Falls was considered historically. The environmental and tourism arguments against development are legitimate and unlikely to be revisited. But at 340 cumecs and 11 metres, the falls represent approximately 33 MW of permanently foregone continuous generation — equivalent to a mid-sized wind farm, running forever, from water that is already flowing.
1.4. The Tongariro Power Scheme: the hidden tributary
The Waikato system's generating potential cannot be understood without the Tongariro Power Scheme, which is the upstream feeder that amplifies everything below. The scheme captures rainfall and snowmelt from the volcanic plateau — the flanks of Mounts Ruapehu and Tongariro — and redirects it north into Lake Taupō rather than allowing it to discharge east.
The scheme comprises two principal power stations: Rangipo (120 MW, underground cavern) and Tokaanu (240 MW), together contributing approximately 1,400 GWh annually. The water they generate from does not disappear — it flows into Lake Taupō and eventually through the entire Waikato cascade, generating electricity a second time at each station downstream. This double-use of the same water molecule is the scheme's fundamental efficiency genius. In aggregate, the Tongariro diversions increased the mean flow through Lake Taupō by approximately 20% — equivalent to roughly 57 additional cumecs passing through every Waikato station downstream, permanently and at zero marginal cost. Without the Tongariro scheme, the Waikato cascade would produce roughly one-sixth less electricity from the same dams.
The Tongariro scheme was transferred to Genesis Energy in 1999 when ECNZ was broken up. Mercury Energy received the Waikato cascade. Genesis also received the Huntly Power Station — New Zealand's largest thermal generator, with up to 1,000 MW of coal and gas capacity. This combination — Genesis controlling both the upstream Tongariro inflows into Taupō and the downstream thermal price-setter at Huntly — is the structural conflict of interest that Part VI examines in detail.
Lake Taupō functions as a vast natural storage reservoir. Mercury actively manages the lake level within consented limits of 355.85–357.25 masl — a 1.4-metre operating range representing just 1% of the lake's total volume. That 1% is enormous: Lake Taupō contains 59 billion cubic metres of water, so each centimetre of lake level represents approximately 61 million cubic metres of potential generation.
1.5. Are the flows producing maximum electricity?
Using P = 9.81 × Q × H × η, across all nine stations with combined net head of approximately 241 metres and mean flow of 340 cumecs, at modern efficiency of 0.91:
P = 9.81 × 340 × 241 × 0.91 ≈ 730 MW → ~6,400 GWh/year (modern turbines)
Against actual annual output of approximately 4,050 GWh, even the conservative estimate suggests the system should be producing around 1,350 GWh more electricity per year — without any increase in infrastructure, without more water, and without raising a single dam. This 33% gap between conservative theoretical potential and actual output is attributable to three causes: the 61 metres of uncaptured head between stations (structural), obsolete turbines (correctable), and market dispatch decisions that systematically withhold generation to drive up prices (a policy failure).
1.6. Wasting water to raise prices: the market gaming problem
The most serious charge against the structure of New Zealand's electricity market is this: the wholesale spot market creates positive incentives for hydro generators to withhold water storage rather than generate electricity, because scarcity drives up the clearing price paid on every megawatt-hour they do dispatch. This is not a theoretical concern. It is documented by the Electricity Authority's own monitoring publications.
New Zealand's electricity market clears on a half-hourly spot price determined by the marginal cost of the most expensive generator dispatched in that half-hour. When hydro storage is high, generators bid cheaply, displace thermal, and the spot price falls. When storage is low, generators bid water at higher prices — ostensibly reflecting scarcity value — and thermal generation, with its high fuel costs, sets the clearing price. Every generator dispatched, including cheap hydro already flowing, receives the same elevated clearing price.
The most dramatic illustration came in August 2024: hydro lake levels fell to a six-year winter low, wind generation was simultaneously low, and available gas supply for thermal generation had contracted. Spot prices at some nodes reached $800 per megawatt-hour — ten to fifteen times the typical level of $80–$120/MWh. Maraetai's 27% annual capacity factor looks very different when those scarce megawatt-hours are selling at $800.
The Genesis Pincer: Controlling Both the Tap and the Price Floor
Genesis Energy emerged from the 1999 ECNZ dismemberment holding two apparently unrelated asset classes: the Tongariro Power Scheme (which pumps water into Lake Taupō), and the Huntly Power Station (which burns gas and coal to generate electricity when hydro is insufficient). These two asset classes appear to serve different purposes. In reality, they form two ends of a single price manipulation mechanism — one that ECNZ's unified management had never needed to contemplate, because a single operator of the entire system had no incentive to manufacture scarcity within it.
The mechanism is structurally elegant. Genesis controls the Tongariro diversions — the tap that fills Lake Taupō. By adjusting the volume of water diverted from the volcanic plateau catchments, Genesis can, within its resource consent conditions, modulate the rate at which Taupō's storage increases. This is operationally defensible on any given day. But the commercial consequence of modulating those diversions downward is entirely predictable: Taupō's level rises more slowly; Mercury rationally bids its Waikato water at higher prices to conserve storage; thermal generation — Genesis's thermal generation at Huntly — is dispatched into the gap; Huntly sets the spot price; and that elevated price is paid to every generator dispatched — including both Mercury's hydro and Genesis's own Tongariro stations.
Genesis therefore profits simultaneously from two streams: from Huntly's thermal generation dispatched at the high spot price it effectively helped create, and from its own Tongariro generation, which also receives that elevated clearing price. This is not collusion in the criminal sense. No agreement between Genesis and Mercury is required. The spot market produces the outcome automatically, as a structural consequence of handing one company the upstream flow-control and the downstream price-floor simultaneously. It is not a conspiracy. It is an incentive structure. In markets, incentive structures are destiny.
The Electricity Authority and its predecessors were aware of the 'undesirable trading situation' concept and investigated several instances of suspected strategic bidding. In a major ruling, the Authority found that Meridian's spilling practices constituted undesirable trading and estimated over $80 million in excess market costs. But the Genesis/Tongariro/Huntly triangle was structurally different and harder to prosecute: Meridian was spilling water that could have generated electricity in the same system it operated. Genesis was managing inflows into a lake it did not operate downstream of — a cross-system, cross-asset arrangement that the Electricity Industry Participation Code was arguably never designed to capture.
The deeper irony is that the architects of the 1999 split specifically divided assets to prevent any single company dominating the market. By giving Mercury the Waikato cascade and Genesis the Tongariro scheme plus thermal, they believed they had created competition. What they had actually created was a pincer: Genesis held the upstream water supply and the downstream price floor, with Mercury — and ultimately New Zealand consumers — trapped in between.
The Government's independent electricity market performance review after the 2024 crisis concluded that the market "lacks reliable back-up options during dry years." This framing normalises the problem rather than diagnosing it. New Zealand does not have a dry-year problem. It has a market structure that systematically manufactures the conditions for price spikes.
1.7. Obsolete turbines: Mercury's own evidence
Mercury's upgrade programme is the most direct evidence available that the Waikato stations have been producing less electricity than they should. The company's published schedule reveals plant installed in the 1950s and 1960s delivering measurably suboptimal performance, and that modern replacements will generate significantly more electricity from the same water flows.
Mercury commenced a $90 million upgrade of Karāpiro in 2019, replacing all three original Kaplan turbines. Installed capacity increased from 96 MW to 112.5 MW — a 17.2% gain — and average annual output increased by 32 GWh, explicitly attributed to a 5% efficiency improvement. Five percent across 75 years of operation at ~505 GWh annual output represents over 1,800 GWh of electricity permanently foregone from water that did flow through the turbines.
Mercury's scheduled future upgrades confirm Karāpiro was not the worst case:
Ōhākurī (2029–2032): +10 GWh annually. Against ~400 GWh output, current plant runs ~2.5% below modern efficiency.
Maraetai I (2030–2034): +45 GWh annually — the largest single upgrade gain. These are among the oldest turbines in the system.
Atiamuri (2031–2034): +32 GWh annually. Against ~305 GWh output, this represents approximately 10.5% underperformance — the largest efficiency gap of any planned upgrade. The Ātiamuri turbines have been running for over 65 years.
The cumulative impact of these three pending upgrades plus Karāpiro is 119 GWh of additional annual generation — electricity that the water already flowing through the system should theoretically be producing today. The gentailer model — vertically integrated generators who also retail electricity — profits from high spot prices as much as from high volumes. An inefficient turbine that justifies a higher spot price bid is, under certain market conditions, more profitable than an efficient turbine that generates more units at a lower price. This is the perverse logic at the heart of the privatised electricity market.
1.8. What a reformed Waikato system could produce
The analysis above permits a reasonably precise quantification of the Waikato system's untapped potential:
Turbine efficiency upgrades (committed): Mercury's programme will deliver +119 GWh per year from Ōhākurī, Maraetai I, Ātiamuri, and Karāpiro by 2034.
Efficiency upgrades at remaining unscheduled stations: Applying a conservative 4% efficiency improvement to Whakamaru, Waipapa, and Aratiatia would yield an estimated additional 65 GWh annually.
Dispatch optimisation at Maraetai: Increasing Maraetai's capacity factor from 27% to 40% would add approximately 680 GWh per year. Under a cost-of-service regulatory model this calculation would never arise.
Cascade head optimisation: Better dispatch co-ordination across the lower cascade could recover an estimated 30–50 GWh annually.
Summing the conservative estimates: committed upgrades (119 GWh) + remaining efficiency gains (65 GWh) + improved dispatch at Maraetai (conservative 340 GWh) + head optimisation (40 GWh) = approximately 564 GWh of additional annual generation from the same river. At a wholesale price of approximately $120/MWh, this represents approximately $68 million of additional annual electricity value — from water that is already flowing, through dams that are already built.
New Zealand's total national electricity generation in 2024 was 43,879 GWh. The combined Waikato cascade and Tongariro scheme currently produce approximately 5,450 GWh annually — about 12.4% of national generation. At maximum realistic efficiency, the combined system could produce approximately 6,800–7,000 GWh per year — roughly 15.5–16% of national generation.
Figure 1 — Hydraulic elevation profile: Tongariro & Waikato
The diagram below maps the full hydraulic staircase of the combined Tongariro–Waikato system from its highest catchment point to sea level. The vertical scale is exact and proportional throughout: the height of each coloured block is directly proportional to the head it represents. Green blocks show Tongariro active generation head; blue blocks show Waikato active generation head; orange-red wedges show free-falling inter-station gaps where water descends with no turbines present.
Figure 1: Hydraulic elevation profile — Tongariro Power Scheme (Genesis Energy) and Waikato River Cascade (Mercury Energy). Vertical scale exact and proportional. Orange-red wedges are inter-station free-fall gaps; the 35 m Whakamaru→Maraetai gap (largest dark-red section) represents ~114 MW of continuously uncaptured potential. Lower four stations are hydraulically chained with near-zero inter-station loss. Source: Klaut Media analysis.
2. The Waitaki River
2.1. Introduction: the South Island's largest power system
The Waitaki River drains the Mackenzie Basin — a vast glacial trough flanked by the Southern Alps to the west and the Two Thumb and Hawkdun ranges to the east. Its catchment encompasses three major glacial lakes — Tekapo, Pukaki, and Ōhau — whose combined storage is the most significant hydrological buffer in the Southern Hemisphere outside Antarctica. The river runs 209 kilometres south-east from the head of Lake Benmore to the Pacific Ocean between Timaru and Oamaru.
The hydroelectric development of the Waitaki proceeded in three distinct phases spanning 57 years. The Waitaki Dam itself — built between 1928 and 1934 during the Great Depression, without earthmoving machinery, by pick and shovel — was followed in the 1960s by Benmore and Aviemore dams on the lower river. The most ambitious phase, the Upper Waitaki development, ran from 1968 to 1985, when four new stations, two major dams, and 56 kilometres of canals were constructed to capture water from all three glacial lake catchments. The upper Waitaki development was at the time the largest hydroelectric project of its kind in the world.
The result is a system of eight power stations generating approximately 7,640 GWh annually — around 18% of New Zealand's national electricity supply and, critically, approximately 75% of the country's total hydro storage capacity. Meridian Energy operates six stations: Waitaki, Aviemore, Benmore, and Ōhau A, B, and C. Genesis Energy — following a Government-directed transfer in 2011 — operates the two Tekapo stations at the upstream end of the system. As with the Tongariro/Waikato split, this ownership division has structural consequences the original designers never intended.
2.2. The physics of the Waitaki cascade
The fundamental power equation applies to the Waitaki exactly as it does to the Waikato. The difference is scale: the Waitaki's combined installed capacity of 1,743 MW dwarfs the Waikato's 1,075 MW. The median flow at Kurow is 356 cubic metres per second, but because the Upper Waitaki scheme captures water from three separate glacial catchments via its canal network, the effective combined flow through the lower stations can approach 600–800 cumecs in good snow years.
The system's genius — and its structural vulnerability — is its storage architecture. Lake Pūkaki alone holds approximately 1,767 GWh of stored energy potential, the equivalent of 9,000 grid-scale batteries. The three primary lakes together represent approximately three-quarters of all hydro storage in New Zealand. This means the Waitaki system is not merely a power station — it is New Zealand's national electricity battery, the buffer against which all other generation variability is managed. When the Waitaki system withholds water, the consequences ripple across the entire national grid.
Using the standard efficiency formula with modern turbine efficiency of η=0.91 and effective combined flow of approximately 400 cumecs:
P = 9.81 × 400 × 610 × 0.91 ≈ 2,185 MW → ~8,400 GWh/year theoretical
Against actual annual output of approximately 7,640 GWh, the gap of roughly 760 GWh is attributable to the same three causes: uncaptured elevation between stages, older turbines operating below modern efficiency, and market dispatch decisions that deliberately withhold generation to elevate the spot clearing price. The Lower Waitaki below the dam — approximately 200 metres of elevation across the braided riverbed from the dam to the sea — represents the system's largest single structural gap, one that Meridian has twice attempted to capture through major infrastructure proposals, and twice failed.
Table 2A — Waitaki hydraulic staircase
Stage / Station | Source level (masl) | Tailrace/outlet (masl) | Net head (m) | Capacity (MW) | Annual output (GWh) | Operator & notes |
Lake Tekapo (source) | 710 | ~690 | ~20 |
|
| Glacial lake; Godley + Macaulay R. inflows |
Tekapo A | 710 | ~690 | ~20 | 25 | 160 | Genesis Energy | 1951 | refurb 2021 ($26.5M seismic) |
Tekapo Canal (26 km) | ~690 | ~520 | ~170 (canal drop) | — | — | Genesis | No generation — entire 170 m elevation drop captured at Tekapo B |
Tekapo B | ~520 | 520 (Pūkaki) | ~170 | 160 | 800 | Genesis Energy | 1977 | feeds Lake Pūkaki |
Lake Pūkaki | ~520 | ~520 | — |
|
| Meridian/Genesis interface. Raised 37 m in 1970s. 1,767 GWh storage — 75% of NZ total |
Pukaki Canal | ~520 | ~520 | ~0 (level canal) | — | — | Meridian | carries Pukaki water to Ohau A |
Lake Ōhau | ~520 | ~520 | — |
|
| Min consent 519.45 m |
Ōhau A | ~520 | ~483 (Ruataniwha) | ~37 | 264 | 1,150 | Meridian | 1980 | canal-fed; into Lake Ruataniwha |
Lake Ruataniwha | ~483 | ~483 | — |
|
| Meridian | formed by Upper Waitaki construction |
Ōhau B | ~483 | ~358 (Benmore) | ~122 | 212 | 970 | Meridian | 1984 | 4 × 53 MW units | no upgrade announced ⚠ |
Ōhau C | ~483 | ~358 (Benmore) | ~122 | 212 | 970 | Meridian | 1985 | completes scheme | no upgrade announced ⚠ |
Lake Benmore | ~358 | ~358 | — |
|
| 1.25 billion m³ | NZ's largest dam | HVDC link southern terminus |
Benmore | ~358 | ~268 (Aviemore) | ~90 | 540 | ~2,200 | Meridian | 1965 | refurb 2008–10 ($67M, +70 GWh). NZ's 2nd largest station |
Lake Aviemore | ~267 | ~267 | — |
|
| Consent range 265.5–268.3 m | largest rotors in NZ (8m dia, 210 t) |
Aviemore | ~267 | ~229 (Waitaki) | ~38 | 220 | ~1,100 | Meridian | 1968 | mid-Waitaki chain |
Lake Waitaki | ~229 | ~229 | — |
|
| Range 227–230.8 m |
Waitaki Dam | ~229 | ~200 | ~29 | 105 | ~420 | Meridian | 1936 | art deco heritage | NZ's first dam not diverting river course |
Lower Waitaki R. (uncaptured) | ~200 | 0 (sea) | ~200 ▲ GAP | — | — | Project Aqua (cancelled 2004) ~520 MW. North Bank Tunnel pending |
TOTALS | 710→0 m | — | Gross: 710 m | 1,743 MW | ~7,640 GWh | 18% national | 75% of NZ hydro storage |
Table 2A — Waitaki hydraulic staircase. The canal system (blue shading) transports water at controlled elevation; all canal elevation drop is captured at adjacent power stations. The purple-shaded lower river gap (~200 m) was the site of the cancelled Project Aqua and ongoing North Bank Tunnel proposals. ⚠ Ōhau B and C flagged as operating with de-rated equipment and no upgrade scheduled. Sources: Meridian Energy; Wikipedia; NZ Geographic; MBIE.
Figure 2 — Waitaki hydraulic elevation profile
The diagram below shows the full hydraulic staircase of the Waitaki scheme from Lake Tekapo (710 masl) to sea level. Green blocks show the Genesis-operated Tekapo stages; blue blocks show Meridian's cascade. The Tekapo Canal conveys water at declining elevation from Tekapo A's tailrace down to Tekapo B. The large purple shading represents the lower Waitaki riverbed — approximately 200 metres of uncaptured head across 60 km of braided riverbed, the intended site of Project Aqua.
Figure 2: Hydraulic elevation profile — Waitaki Hydro Scheme. Green = Genesis Energy (Tekapo A & B). Blue = Meridian Energy (Ōhau A/B/C, Benmore, Aviemore, Waitaki Dam). The 2011 Government-directed split gave Genesis the upstream Tekapo tap while Meridian retained all downstream infrastructure. Purple zone shows uncaptured lower river (~200 m). Vertical scale exact and proportional. Source: Klaut Media analysis.
2.3. The 2011 Tekapo split: replicating the Genesis Pincer
The ownership structure of the Waitaki scheme replicates, with disturbing precision, the structural conflict of interest described in the Waikato analysis. When the New Zealand Electricity Corporation was broken up in 1999, the Waitaki cascade went to Meridian Energy and the Tekapo stations remained with Meridian. In 2011, the Government directed a further structural division: Tekapo A and B were transferred to Genesis Energy, the same company that controls the Tongariro scheme upstream of the Waikato.
The parallel is exact. Genesis now controls the upstream water tap — the rate at which Tekapo water flows through the canal system into Lake Pūkaki and thence to all of Meridian's downstream stations — while Meridian manages the downstream storage and dispatch. Neither is accountable for the system's total energy output. And Lake Pūkaki — which holds 1,767 GWh of stored energy, the equivalent of 9,000 of the grid-scale batteries being celebrated by the industry — is the largest single piece of leverage over national electricity prices that any New Zealand generator controls.
The clearest evidence that this structure produces perverse market outcomes was the Electricity Authority's landmark ruling in 2020. The Authority found that Meridian had engaged in 'undesirable trading practices' — specifically, spilling water from its Waitaki lakes without generating electricity, in circumstances where the spilling benefited Meridian's retail and hedge book position. The ruling estimated market costs to consumers in excess of $80 million. Water was physically disposed of — rather than generating electricity — because the commercial structure of the spot market made disposing of it more profitable than running it through the turbines. This is the reductio ad absurdum of the privatised electricity market: the country's largest renewable energy asset, built by taxpayers, operated by a majority Crown-owned company, disposing of free water because the market made disposal more profitable than generation.
2.4. Benmore's efficiency gap and the cost of deferred upgrades
Benmore Power Station is, after Manapōuri, the second-largest hydroelectric station in New Zealand. Its six Francis turbines — commissioned between 1965 and 1966 and operating until 2008 on their original runners — were shown to have declining efficiency between 1965 and 2002, with a measurable shift in peak efficiency load as the turbine runners wore. The 2008–2010 refurbishment at a cost of approximately $67 million achieved a 6% reduction in water use per unit of generation, yielding an additional 70 GWh of annual output.
At 2,200 GWh annual output over 43 years from commissioning to refurbishment, the 6% efficiency deficit represents approximately 5,660 GWh of permanently foregone generation — electricity that was lost to turbine wear and the market's disincentive to invest in efficiency when scarcity-based pricing offers an alternative profit mechanism. At $120/MWh, that is approximately $680 million of electricity that never flowed to consumers, from water that flowed freely through the turbines but was converted at suboptimal efficiency.
The Ōhau stations — B and C, both commissioned in the early 1980s — have operated for over 40 years on their original turbines. The MBIE hydro generation stack report notes that Ōhau B and Ōhau C are among the stations flagged as operating with de-rated equipment and measurable inefficiencies. No upgrade programme for either station has been publicly announced. At 970 GWh each per year, even a 3% efficiency deficit across 40 years represents approximately 1,165 GWh per station of permanently lost generation from the same water.
2.5. Project Aqua and the road not taken
The most significant structural underperformance in the Waitaki scheme is the 200 metres of uncaptured elevation in the 60-kilometre braided reach of the lower Waitaki River below the Waitaki Dam. The river exits the dam at approximately 200 masl and falls gradually through the Canterbury Plains to sea level, losing potential energy along a reach that was deliberately left undeveloped.
In 2001, Meridian Energy lodged Project Aqua — a proposal for six dams on a man-made canal running from the Waitaki Dam to the sea, capable of generating approximately 520 MW from water already flowing through the existing stations above. By Meridian's own 2004 annual report, the company had spent $38.7 million on design and consenting before abandoning the proposal on 29 March 2004. Public opposition, RMA complexity, and land access negotiations were the stated reasons, though flat electricity demand forecasts also weakened the commercial case.
The successor proposal — the North Bank Tunnel — received water rights in 2009 but saw land access negotiations suspended in 2013 when flat demand forecasts again made the project uneconomic. As of 2026, the lower Waitaki remains undeveloped — 200 metres of head, 60 kilometres of river, and approximately 520 MW of potential capacity flowing freely to the sea. The cancellation of Project Aqua is the most explicit single example in New Zealand's electricity history of the market failing to capture publicly available renewable energy potential: a scheme whose feasibility was demonstrated, whose costs were substantially incurred, and which was abandoned because the spot market did not provide a commercial signal sufficient to justify investment in firm renewable generation.
2.6. Waitaki: national share and quantified potential
The Waitaki scheme at current output produces approximately 7,640 GWh annually — 17.4% of the 2024 national total. Maximum realistic output — Ōhau B and C at full efficiency with updated turbines, Benmore and Aviemore maintained at post-2010 efficiency, and dispatch optimised system-wide — could plausibly reach 8,200–8,500 GWh annually: a gain of 560–860 GWh from the same water and the same infrastructure.
These are conservative estimates. They exclude the lower Waitaki gap entirely. Even within the existing infrastructure, the identified gains are substantial: approximately 70 GWh/year from Ōhau B and C turbine efficiency improvements; a further 40–80 GWh from eliminating market-driven withholding at Benmore; and 50–100 GWh from optimised cascade coordination between the Genesis Tekapo stages and Meridian's downstream infrastructure. At $120/MWh, the conservative identified gain represents approximately $67–103 million per year in consumer benefit from existing infrastructure.
3. The Clutha / Mata-Au River
3.1. Introduction: New Zealand's largest river by volume
The Clutha / Mata-Au is New Zealand's largest river by volume. Its mean discharge of approximately 500 cubic metres per second at Balclutha exceeds any other river in the country, fed by the combined catchments of three major lakes — Wānaka, Hāwea, and Wakatipu — whose glaciated mountains receive up to 8,000 mm of annual precipitation on the western ranges. The river runs 340 kilometres from the outflow of Lake Wānaka at Alberttown to the Pacific Ocean at Balclutha, descending 278 metres of elevation through the schist gorges and alluvial plains of Central Otago.
Despite this extraordinary hydrological resource — the largest flow of any New Zealand river, draining the most heavily glaciated catchments in the country — the Clutha is spanned by precisely two dams. Clyde Dam (432–464 MW, commissioned 1992) and Roxburgh Dam (320 MW, commissioned 1956) are separated by approximately 60 kilometres of river and together produce approximately 3,700 GWh annually — about 8.4% of New Zealand's national generation. Both are owned and operated by Contact Energy, which unlike Mercury or Meridian is not majority Crown-owned but is a publicly listed private company.
The Clutha system's underperformance is not primarily a story of market gaming, obsolete turbines, or inter-station gaps — though all three exist. It is, above all, a story of radical underdevelopment: a river producing 500 cumecs and 278 metres of gross head being captured by only two dams, generating a small fraction of its theoretical potential, with the remainder flowing freely to the sea. The upper reach from Lake Wānaka to Clyde's headpond drops 46 metres over approximately 80 kilometres of free river — uncaptured. The lower reach from Roxburgh's tailrace to the sea descends approximately 112 metres over 160 kilometres — also uncaptured. Together, these two uncaptured reaches represent more than half the river's total head and a very large share of its latent generation potential.
3.2. The physics of the Clutha
At mean flow of 500 cumecs and gross head of 278 metres, the Clutha's theoretical maximum generation potential at modern turbine efficiency (η=0.91) is:
P = 9.81 × 500 × 278 × 0.91 ≈ 1,245 MW → ~10,900 GWh/year theoretical
Against actual annual output of approximately 3,700 GWh, the gap between theoretical potential and actual generation is approximately 7,200 GWh — roughly 66% of the river's physical capacity flowing to the sea as unused energy. The structural reasons are: a 46-metre free river gap from Wānaka to Clyde's headpond; a 100-metre drop captured at Clyde (well-utilised); a 20-metre drop at Roxburgh (partially captured); approximately 112 metres of lower river descent completely uncaptured; and 70 metres of head across the Hāwea dam that generates nothing, built purely as a storage control structure.
Table 3A — Clutha hydraulic staircase
Stage / Station | Headpond (masl) | Tailrace/outlet (masl) | Net head (m) | Capacity (MW) | Annual output (GWh) | Notes |
Lake Wānaka (source) | ~278 | ~278 | — |
|
| Natural outlet, no dam. 192 km². Clutha source. |
Lake Hāwea (storage dam) | ~348 | ~278 (Hāwea R.) | 70 (no turbine) ▲ | — | — | Raised 18 m by 1958. No power station — pure storage. 70 m head entirely wasted. Fast-Track proposal: raise +2 m, still no turbine |
Upper Clutha free river (Wānaka→Clyde) | ~278 | ~232 (Lake Dunstan) | 46 m ▲ GAP | — | — | ~180 MW uncaptured. Kawarau River (Wakatipu) joins at Cromwell. Long-recognised viable dam site. |
Clyde Dam | ~232 | ~132 (Lake Roxburgh) | 100 | 464 | ~1,900 | Contact Energy | 1992 | 102 m concrete gravity dam | geological fault — cut from 612 MW, 45% cost overrun |
Lake Roxburgh | ~132.6 | ~132 | — |
|
| 290 GWh storage. Sediment trap: Alexandra riverbed +3.6 m since 1956. |
Roxburgh Dam | ~132 | ~112 | ~20 | 320 | 1,400–1,830 | Contact Energy | 1956 | 8 Kaplan turbines | First upgrade 2024–26: 4 turbines replaced, +44 GWh. Remaining 4 on 1956–61 runners. |
Lower Clutha (Roxburgh→sea) | ~112 | 0 (sea) | ~112 ▲ LARGE GAP | — | — | ~160 km to Balclutha. ~5 dam sites previously identified. All proposals abandoned. Ecologically significant braided river habitat. |
TOTALS | 278→0 m | — | Gross: 278 m | 784 MW | ~3,700 GWh | ~8.4% national. 228 m of gross head (82%) uncaptured. Only 2 dams on 340 km river. |
Table 3A — Clutha / Mata-Au Hydraulic Staircase. Purple shading indicates uncaptured reaches. Lake Hāwea is a unique case: 70 metres of head across an existing dam, with no power station ever installed. Sources: Wikipedia; Contact Energy; Otago Regional Council; NZ Geographic.
Figure 3 — Clutha / Mata-Au hydraulic elevation profile
The diagram below shows the full elevation profile of the Clutha from Lake Wānaka (278 masl) and Lake Hāwea (348 masl) to sea level. The dark red wedge represents the 46-metre upper Clutha free-river gap. Clyde captures the largest single block of head (100 m). Roxburgh captures a shallow 20-metre drop. The purple zone shows approximately 112 metres of lower Clutha descending to sea level across 160 km of uncaptured river.
Figure 3: Hydraulic elevation profile — Clutha / Mata-Au scheme, Contact Energy. Dark red = uncaptured upper river gap (46 m). Blue = active generation. Purple = uncaptured lower river (~112 m across 160 km). Lake Hāwea (leftmost block, 348 m) has 70 m head across its control dam with no power station. Vertical scale exact and proportional. Source: Klaut Media analysis.
3.3. The Roxburgh scandal: 68 years without a turbine upgrade
Roxburgh Dam's turbines were commissioned in 1956 and 1960. They operated for 68 years — more than three generations — without a single turbine replacement. Contact Energy's 2024 announcement that it would replace four of the eight original turbines at a cost of $30 million, with the first new turbine operational in January 2025 and the programme completing in 2026, is welcome. The 44 GWh gain expected from the four replacements alone represents a measurable efficiency deficit in the existing turbines — equivalent to approximately 3% underperformance against modern design.
The 68-year deferral is more striking than the 75-year Karāpiro deferral documented in the Waikato analysis. Roxburgh's turbines are Kaplan units — an axial-flow design particularly sensitive to runner wear in high-sediment environments. The Clutha is not a low-sediment river. The combination of the Shotover River's glacial silt load (before Clyde Dam intercepted it in 1992) and the Manuherikia River's sediment contribution has elevated the Alexandra-Roxburgh riverbed by 3.6 metres since the lake formed. That same sediment has flowed through Roxburgh's turbine runners continuously for 68 years.
At Roxburgh's mean annual output of approximately 1,615 GWh, a 3% efficiency deficit over 68 years represents approximately 3,290 GWh of permanently foregone generation. At $120/MWh, that is approximately $395 million of electricity that was not generated from water that did flow through the turbines. The decision not to upgrade for 68 years reflects exactly the same market logic identified at Karāpiro, Ātiamuri, and Benmore: the spot market rewards operators who can bid water at scarcity prices rather than maximising generation from every available cubic metre. A turbine that extracts 97% of the available energy per cubic metre is, in some market conditions, less profitable than a turbine that extracts 94% but can be bid into the market at the moment of peak scarcity price.
The remaining four turbines — not included in the current upgrade programme — continue to operate on their original 1956 runners. Contact Energy has not announced a second phase. When the first four upgraded turbines are running, half of Roxburgh will be operating at modern efficiency and half at 1950s efficiency. This is not satisfactory for infrastructure that supplies electricity to more than 250,000 homes and represents 8% of New Zealand's South Island hydroelectric generation.
3.4. The Contact Energy conflict: a third pincer
Contact Energy is the only major New Zealand generator that is neither majority Crown-owned nor a cooperative. It is a publicly listed company, responsible to its shareholders, and its obligations to the public interest are limited to those imposed by the Resource Management Act, the Electricity Industry Participation Code, and its resource consents. This matters because Contact holds both the Clyde and Roxburgh dams and the Taranaki Combined Cycle gas station (approximately 550 MW), one of New Zealand's largest gas-fired thermal generators.
The structural conflict of interest is identical in form to the Genesis pincer documented in the Waikato analysis. Contact can modulate the release of water from Lake Hāwea within its resource consent conditions. When Hāwea's level falls, Clyde and Roxburgh generate less; when thermal generation must fill the gap, Contact's Taranaki Combined Cycle dispatches at the clearing spot price — a price that is, in part, a function of the hydro scarcity Contact itself has helped create. No collusion with any other generator is required. The spot market produces the outcome automatically.
Contact's current Fast-Track Approvals Act application to raise Lake Hāwea by two metres will increase the scheme's storage capacity and thus its strategic value for exactly this kind of pricing leverage. It is presented as a measure to support grid security during dry periods. In a market that rewards scarcity, it is simultaneously an expansion of Contact's strategic resource for managing dry-period spot prices.
3.5. Sediment: the existential threat
The Clutha faces a structural challenge that the Waikato and Waitaki do not: sediment. The Shotover River was the primary feeder of suspended sediment to the pre-dam Clutha. When Roxburgh Dam was built in 1956, that sediment began to accumulate in Lake Roxburgh. By 1979, the riverbed downstream of the Alexandra bridge had risen 3.6 metres above its pre-dam level. The 1999 flood, only 80% the volume of the 1878 flood, nonetheless caused higher water levels in Alexandra because of the elevated riverbed.
The construction of Clyde Dam in 1992 intercepted the Shotover's sediment inflow into Lake Roxburgh, creating a new sediment trap in Lake Dunstan. By December 1999, 5.7 million cubic metres of sediment had accumulated in the arm of Lake Dunstan fed by the Shotover, with the 1999 flood alone contributing 2.8 million cubic metres in a single event. Lake Dunstan's total volume provides an estimated 100+ years of sediment storage capacity — but as that capacity fills, the options for managing it without environmental impact to downstream communities narrow considerably.
The combination of rising riverbeds, flood elevation, and sediment management creates direct costs for Contact Energy and for downstream communities — costs that do not appear in the electricity market's half-hourly price signals. The sediment issue is a material constraint on the Clutha scheme's long-term operation that no market mechanism is designed to price.
3.6. Clutha: national share and quantified potential
The Clutha scheme at current output produces approximately 3,700 GWh annually — 8.4% of the 2024 national total. At maximum realistic output — Roxburgh with all eight turbines upgraded to modern efficiency, Clyde operating at its full 464 MW capacity, Lake Hāwea managed optimally, and dispatch decisions made without reference to Contact's thermal portfolio — the combined scheme could produce approximately 4,200–4,500 GWh per year: a gain of 500–800 GWh from the same water and infrastructure.
The Roxburgh turbine upgrade (four turbines, +44 GWh) is the most direct gain currently being realised. The remaining four turbines — still operating on their 1960 runners — represent a further estimated 30–40 GWh of recoverable efficiency, for a total post-upgrade gain at Roxburgh alone of approximately 74–84 GWh annually. Clyde's efficiency is less well-documented publicly, but its geological complications and the absence of any published upgrade programme suggest it may be operating below what its Francis turbine design could achieve with modern runner profiles. A conservative 3% efficiency improvement at Clyde would yield approximately 57 GWh per year.
The structural gaps — 46 metres of upper Clutha free river, 70 metres of Hāwea head with no turbine, and 112 metres of lower Clutha to the sea — represent the scheme's largest potential gains but also its largest engineering and political challenges. The lower Clutha gap (160 km, 112 m head, 500 cumecs) alone has a theoretical generation potential of approximately 2,700 GWh per year — more than the entire current Clutha output. Whether environmental, cultural, and ecological considerations are compatible with that level of development is a policy question, not a physics one.
4. The combined picture
4.1. Three rivers: a national electricity audit
The three river systems analysed in this report — Waikato/Tongariro, Waitaki, and Clutha/Mata-Au — together account for approximately 38% of New Zealand's national electricity generation. The following table summarises their combined performance against theoretical potential:
Table 4A — Combined Performance Audit: All Three Systems
System | Operator(s) | Current output (GWh) | % national | Realistic max (GWh) | Key constraints |
Waikato + Tongariro | Mercury + Genesis | ~5,450 | 12.4% | ~7,000 | Genesis pincer (Tongariro/Huntly); Maraetai 27% CF; obsolete turbines (upgrade programme to 2034); 61 m inter-station gaps |
Waitaki | Meridian + Genesis | ~7,640 | 17.4% | ~8,400 | 2011 Tekapo split; undesirable trading ruling ($80M+); Ōhau B/C turbines (40+ yrs, no upgrade); lower river uncaptured (Project Aqua site, ~520 MW) |
Clutha / Mata-Au | Contact Energy | ~3,700 | 8.4% | ~4,500 | 68-yr turbine deferral (3,290 GWh foregone); Contact thermal/hydro conflict; Hāwea 70 m head wasted; upper/lower river uncaptured; sediment management |
COMBINED | — | ~16,790 GWh | ~38.3% | ~19,900 GWh | ~3,100 GWh/year recoverable within existing infrastructure — without new dams, new water rights, or major new consents |
Table 4A — Combined performance audit. Current and realistic maximum outputs compared. Realistic maximum assumes modern turbine efficiency throughout, optimised dispatch without market withholding, and no structural changes to uncaptured head reaches. Figures exclude structural gaps requiring new infrastructure. Source: Klaut Media analysis; Mercury Energy; Meridian Energy; Contact Energy; MBIE.
The aggregate picture is striking. From three river systems accounting for more than a third of national generation, approximately 3,100 GWh per year of additional renewable electricity is theoretically recoverable without building a single new dam, acquiring a single new water right, or significantly altering the environmental footprint of any of these schemes. At 43,879 GWh national consumption (2024), this 3,100 GWh represents 7% of national generation — enough to eliminate nearly all coal-fired electricity in a drought year, or to power approximately 370,000 additional average New Zealand homes.
The price effect — as argued throughout this report — is disproportionately large relative to the volume increase. Each of these systems is a marginal price-setter in the half-hours that dominate annual average spot prices. Benmore is the HVDC link's southern terminus; the marginal spot price at Benmore in a dry winter sets the price floor for most of the national grid. Maraetai's capacity factor affects Mercury's Taupō storage decisions. Roxburgh's dispatch shapes Contact's Taranaki Combined Cycle bidding strategy. Optimising all three simultaneously would eliminate the extreme price events that account for a disproportionate share of annual consumer electricity costs.
The 2023 and 2024 data illustrate the causal chain with unusual clarity. In 2023, hydro generation reached its highest level since 2004, renewables reached 88.1% of national generation — their highest share since 1981 — and coal generation fell 17.6%. Prices were benign. In 2024, dry conditions reduced hydro generation 11% below the prior year and coal generation increased 118% to 2,243 GWh to compensate. Wholesale prices rose sharply from July, with some nodes briefly touching $800/MWh during August. The causal relationship between available hydro and market price is not theoretical — it is measurable year by year in MBIE's own published data.
4.2. Conclusion: the privatisation dividend that never was
The 1999 restructuring of New Zealand's electricity sector was premised on a specific theory: that competitive generators, each owning a subset of generation assets, would have stronger incentives to maximise renewable output than a vertically integrated state monopoly. Thirty years of evidence from the Waikato, Waitaki, and Clutha systems suggests the opposite: the competitive structure has systematically weakened the incentive to maximise renewable generation, by creating market conditions in which withheld water is more valuable than dispatched water, in which deferred turbine investment is more profitable than efficient turbine investment, and in which cross-asset conflicts of interest reward managing renewable generation in a way that benefits thermal assets.
The Genesis pincer — Tongariro inflows paired with Huntly dispatch; the 2011 Tekapo split pairing upstream Tekapo inflows with Genesis's South Island portfolio — was not designed. It emerged from the incentive structure of the market. The Meridian undesirable trading ruling — water physically spilled into the Waitaki River rather than converted to electricity, because the market made spilling profitable — was not designed. It too emerged from the incentive structure of the market. The 68-year deferral of Roxburgh's first turbine upgrade was not designed. It emerged from the market's failure to provide a commercial signal sufficient to justify investment in renewable efficiency when the alternative was bidding existing capacity at scarcity prices.
New Zealand's electricity debate in 2026 centres on new capacity: wind farms, solar, grid-scale batteries. These are important investments. But the analysis in this report suggests that the more immediate opportunity — and the one that requires the least capital, the least new infrastructure, and the least environmental impact — is to capture the renewable generation that is already flowing through rivers whose dams were built by previous generations, whose physics have not changed, and whose market managers have been given rules that systematically reward the management of scarcity over the maximisation of abundance.
The Waitaki River, as Meridian's own promotional material puts it, is the heart of New Zealand's electricity system. The Clutha/Mata-Au is New Zealand's largest river by volume. The Waikato cascade was the first major hydroelectric system built in New Zealand, and remains the largest single generation asset in the North Island. Together, they are producing substantially less electricity than the physics permits.
The question is not whether New Zealand could generate more clean energy from these rivers. The question is whether New Zealand's political economy has the will to change the rules under which they are managed.
References and Sources
Wikipedia — Waikato River (flow, cascade overview, 13% national generation)
Mercury Energy — Karāpiro Hydro Station (58 m dam height, 112.5 MW upgrade, +32 GWh)
Mercury Energy — Lake Levels (355.85–357.25 masl operating range)
Mercury Energy — Taupō Control Gates (flow management, cascade description)
Wikipedia — List of dams and reservoirs in New Zealand (installed capacities)
Wikipedia — Karāpiro Power Station (58 m height, 148 cumecs minimum flow)
Wikipedia — Aratiatia Power Station (run-of-river, 55 ha headpond)
Wikipedia — Maraetai Power Station (construction history, arch dam, 27% CF)
Wikipedia — Whakamaru Dam (diversion channel 31 m deep, grid switching station)
Wikipedia — Tongariro Power Scheme (1,400 GWh, Genesis Energy)
Wikipedia — Electricity sector in New Zealand (Genesis/Huntly, 1999 split)
Engineering New Zealand — Arapuni Power Station and Dam (heritage record)
Engineering New Zealand — Ātiamuri Power Station (43.6 m concrete dam, Alexander Gibb & Partners)
Electricity Authority — Hydro storage limits and why rain matters for electricity prices
Electricity Authority — High hydro levels and wholesale prices
Electricity Authority — The impact of our climate on hydro generation
Electricity Authority — NZ electricity system in August and November 2024
Electricity Authority — Retail power prices and power supply security
MBIE — Government response to independent electricity market performance review
Russell McVeagh — Energy Sector: Reviewing 2024 and Looking to 2025
Grokipedia — Meridian Energy (Electricity Authority undesirable trading ruling, $80M+ market costs)
Stuff / Waikato Times — SOE cited for Taupo erosion (resource consent review)
MBIE — Energy in New Zealand 2024 (hydro generation 26,309 GWh; renewables 88.1%; coal -17.6%)
Meridian Energy — About the Waitaki Hydro Scheme (history, 56 km canals, 8 stations)
Meridian Energy — Lake Levels (Benmore 355.25–361.45 m; Aviemore 265.5–268.3 m)
Wikipedia — Waitaki River (station list, cumecs, catchment description)
Wikipedia — Benmore Dam (540 MW; 2nd largest; 1.25B m³; HVDC; $67M refurb 2008–10 +70 GWh)
Wikipedia — Waitaki Dam (built without machinery; 1936; Depression era)
Wikipedia — Lake Tekapo (710 masl; 2011 transfer to Genesis; Tekapo A 1951; $26.5M seismic 2021)
Wikipedia — Clutha River (500 m³/s mean flow; 340 km; catchment description)
Wikipedia — Clyde Dam (464 MW; 4 Francis turbines; 102 m height; geological fault; 45% cost overrun)
Wikipedia — Roxburgh Dam (1956; 8 Kaplan turbines; sediment; Alexandra riverbed +3.6 m)
Contact Energy — New Turbines at Roxburgh Dam (Voith Hydro; +44 GWh; $30M; first turbine Jan 2025)
Contact Energy — Clutha Hydro Scheme (Fast-Track application; Lake Hāwea +2 m)
Wikipedia — Lake Dunstan (Clutha; Clyde 1992; Cromwell Gorge)
Grokipedia — Clyde Dam (432 MW; Francis turbines; geological issues)
NZ Geographic — What Price a River? (Clutha scheme history; Hāwea raised 18 m)
NZ Geotechnical Society — Geology and the Clyde Dam (fault discovery; grouting; seismic hazard)
Otago Regional Council — Clutha/Mata-Au FMU (resource consent structure; environmental flows)
RNZ — ORC to Review Clyde Dam Resource Consent (Kawarau Arm; Contact Energy; environmental impact)
