Lake Onslow pumped storage

Comment by Earl Bardsley, University of Waikato (formerly Associate Professor in hydrology and Meridian Energy Senior Research Fellow in Applied Hydrology).

October 2025

Figure 1: The Lake Onslow reservoir on February 21, 2023. Energy storage would be achieved by raising the reservoir to about 80 metres above its present level.

There has been an interesting recent development concerning constructing the Lake Onslow pumped storage scheme by a private consortium (Pumped Hydro Holdings). The directors are former Transpower chair Keith Turner, former Environment Minister David Parker, former Meridian renewable energy generation manager Ken Smales, and senior environmental and commercial lawyer John Hardie. Current shareholders include the four directors of the consortium and a company owned by businessman and former merchant banker Rodger Finlay. Finlay is the current acting chair of the Reserve Bank and a former chair of NZ Oil and Gas – now Echelon Resources.
The intention is to lodge an application for fast-track on the project, for 5 TWh of energy storage capacity and 1000 MW of pumping/generating capacity. Turner says that there is strong interest from potential investors, with financing of the project being extremely appealing to very large sovereign wealth funds and investors.
Last month, Resources and Associate Energy Minister Shane Jones told Energy News that he would welcome a move by private investors to get the Onslow project under way.

Background to the current need for the Lake Onslow scheme for electricity prices, and the futility of imported LNG.

The October 2025 market review report (Frontier Economics) emphasises the critical economic importance of avoiding high prices in hydro dry years. Energy Minister Simon Watts notes in a press release that it can take up to 25 years for the economy to recover dry year high power prices. (And the average time between dry years is much less than 25 years).

In response to the October 2025 release of the Frontier report, the government is considering funding facilities that will enable importing LNG for provision of additional power in dry years. However, this will only be of value for putting a cap on electricity prices at a high value somewhere greater that $200 / MWh. That is, it is not be economical to generate electricity from LNG fuel unless the wholesale price exceeds at least $200 / MWh. While this is an improvement over the prices that went up to $800 / MWh in the 2024 dry winter, having $200 / MWh or more is still not going to prevent industry closures from an inability to compete on the world market. Also, taking its full production path into account, burning imported LNG for electricity generation produces as much carbon emission as coal per unit of power generated.

A better alternative to LNG for dry years in the short/medium term is to define additional “fuel” as the deeper water of Lake Pukaki, providing the additional electricity from the Waitaki hydro power stations. Meridian is already taking steps to apply for a lowered minimum operating level of Lake Pukaki.

However, the massive visual impact of a drawn-down iconic scenic lake is not something that should be accepted as a permanent means of keeping electricity prices down in dry years.

Another factor is that the multi-fuel Huntly power station will be coming to the end of its life around 2035.

One attractive solution is to further develop geothermal power to play the main baseload power role by 2025, coupled with pumped storage at Lake Onslow (and perhaps at other locations as well) coming in as Huntly closes. This will ensure moderate power prices, given the high-price contribution of dry-year fossil fuels is largely removed.

It happens that the Lake Onslow basin in Otago is unique in that it can store a large volume of water, pumped up to it from the Clutha River far below. Its combination of water volume and elevation gives an energy storage equivalent to having a second Lake Taupo, plus as second Lake Pukaki, plus a second Lake Manapouri, and so on to duplicate every hydro lake in New Zealand – all at a single site.

Combined with a generating capacity more than the Huntly power station, the Lake Onslow scheme is perfectly suited to provide the low-cost flexible firming that is needed to keep power prices down as wind and solar generation expands. More than that, the Lake Onslow scheme being in the infrastructure pipeline will unleash a significant build of new wind and solar generation capacity that would not otherwise happen. See also under the heading The scheme would encourage additional renewable generation further down.

The reason for the low cost of Onslow flexible firming electricity is that the “fuel” for the Lake Onslow scheme is cheap renewable power that provides the energy for pumping at times of abundant wind/solar/hydro. That is, Onslow power is just recycled renewable power without the intermittency issue. Pumping costs would still have to be recovered in the price of course. However, Onslow flexible firming electricity would still be far cheaper than that from burning coal / fossil gas / diesel / bomass.

The Onslow scheme would have a secondary cost-reduction effect as well. Just the presence of a new lake at elevation with significant water storage will have the effect of reducing prices. That effect is seen already with the existing hydro lakes – when their levels are high, wholesale power prices tend to be low.

The strong opposition to the Lake Onslow scheme by the gentailers is understandable – having a new entity offering cheaper electricity into the market is not in their financial interests.

It still remains uncertain whether the Lake Onslow scheme will go ahead or not. But its presence or absence will determine the nature of New Zealand’s energy scene (and the economy) for most of the next hundred years.

The remainder of this post lists some specific features of the Lake Onslow scheme in more detail, which may be of interest to some.


The Lake Onslow scheme would be in near-continuous operation

Although the Onslow scheme’s main motivation is for cheaper (and emission-free) power in dry years, there is no such thing as a “reserve” pumped storage scheme in any electricity market. Pumped storage operates by pumping when prices are low, doing nothing when prices are intermediate, and generating when prices are high. Generation is triggered when prices rise high enough to recover pumping cost and also allow some net income.

In this way, the Onslow scheme would generate some income even during the initial filling of an expanded Lake Onslow to its mean operating level. This is because the lake would fill as a sawtooth rising trend, incorporating switching from pumping to periods of generation when prices were sufficiently high.

It would result in a significant reduction or total removal of fossil fuels for electricity generation.

Onslow, when generating, would largely eliminate fossil fuel power generation because it would offer cheaper electricity into the market. Imported coal and LNG are expensive fuels, subject to world market price fluctuations.

Substitutions of fossil fuel power by Onslow electricity would occur frequently through the year. As shown in Fig. 2, coal has been used consistently from one year to the next (including 2025 from mid-January), though with greater use in the 2021 and 2024 dry years.

Figure 2. Cumulative mass of coal used to generate electricity, from 2016. Source: MBIE.

Onslow operation in itself would not, however, cover all eventualities. For example, it would not be possible to guarantee that every North Island winter peak demand could be met by northward transmission of Onslow power through the grid. Some limited gas for peaking is likely to be required in the first instance, perhaps to be replaced later by batteries or small North Island pumped storage schemes. Also, every thirty years or so there will be times of consecutive dry periods when there is need to resort to deep hydro lake water.

At $16b, the Lake Onslow scheme may come at no cost to the nation.

The purpose of a power scheme is of course to produce new electricity, added to what was there before. In contrast, Onslow generation produces replacement electricity. That is, when generating, Onslow power would substitute for electricity that would otherwise have been derived from burning fossil fuels, particularly when hydro inflows are low. The substitution occurs because Onslow would offer electricity into the market before the price rose high enough to make fossil fuel use economical. This capping of prices leads to cheaper power.

The scheme’s construction cost could, for example, be paid for over time as a cost recovery levy on consumer electricity bills. However, the levy would be invisible because it has been estimated that, spread over time, the capital cost of Onslow would be completely offset through the lower electricity prices to consumers (The economics of four future electricity system pathways for New Zealand p.49). Something similar may apply in the event of the Onslow scheme being constructed by the private sector (presently under consideration).

Constructing the Onslow scheme therefore need not divert government funding away from important areas like health and education.

Electricity futures prices would be reduced

Large electricity users often take out insurance against future high electricity prices by negotiating a fixed price contract for some duration. This is similar to a fixed mortgage. The providers of such contracts must take all risk factors into account – insurers do not wish to lose money.

If insurers see there is a high risk of future high electricity prices, they will raise their contract price accordingly. Purchasing high-price hedging contracts can result in energy-intensive products having selling prices that are too high to be competitive in overseas markets.

Futures prices can also remain high long after the high spot prices of a dry year have returned to normal. This is because a recent dry year is seen as indicating a risk that another may occur.

An additional risk factor is that New Zealand power supply security in dry years is heavily dependent on the aging Huntly power station.

A 400 MW Huntly unit failed in June 2023 and was unavailable for some months. If that failure had happened in June 2024, the winter price peaks would have been much higher. A dry year coincident with a Huntly failure is therefore an additional risk factor that will tend to increase electricity futures prices.

The dry year insurance role of the Onslow scheme would be to distribute the dry year risk buffer over multiple spatial locations. Operating continuously in the market, the scheme would result in the main hydro lakes being closer to their mid-range levels more often.

That is, the hydro lake levels will be higher than they otherwise would be at the start of a period of low river inflows to the lakes, so a Huntly failure would be less of a risk.

An economic and climate change advantage

There are economic impacts from complete or partial closures of export industries because of high power prices. The export reductions in the 2024 dry winter have been estimated as causing a $300m loss to the economy, contributing to New Zealand’s worst 6-month GDP (excluding covid times) since 1991 – which incidentally was also a low-inflow year.

An unquantifiable but important additional economic loss factor is overseas industry executives who make silent decisions not to set up in New Zealand because of perceived unreliable and expensive power supply.

The Lake Onslow scheme could carry out multiple energy functions simultaneously

The Onslow scheme would be in generation mode more frequently, on average, in winter and in dry years. However, it would also provide firming at all times of year for large-scale wind energy.

Being able to alternate between pumping and generating means that the 1000 MW scheme could provide the necessary firming for up to 2000 MW of new South Island wind power. There would be no net change in Lake Onslow water level from this wind power support operation, which would superimpose on seasonal and dry year operation.

Also, having Onslow’s on-call spare generating capacity would help system security. On the cold morning of May 10, 2024 there was a public call for reduced power consumption because there was limited available backup generation in the event of a unit failure at some location.

The scheme would encourage additional renewable generation.

In pumping mode, the Onslow scheme would be a purchaser of power and set a floor price for wind and solar power sales. This improves the economics of setting up new generation capacity (The economics of four future electricity system pathways for New Zealand p.32).

Paradoxically, the suggestion was made during the election campaign (and still made by the current government) that the Onslow scheme, just by being under consideration, would have a “chilling effect” that would in fact deter new renewable generation. The argument was that Onslow in generation mode would undercut renewable power in the market, so new renewable projects would be less likely to proceed.

Onslow power would undercut gas and coal. However, it could not compete with wind, solar and hydro generation because the Onslow selling price for its power would have to include pumping cost and a profit margin. Unsurprisingly, there is no evidence of the Onslow scheme having had a chilling effect on renewable developments and an element of gentailer misinformation may have been involved.

Even more importantly, pumped storage at Lake Onslow would unleash an energy ecosystem of new wind and solar generation that otherwise would not happen. This is because wind and solar farms are not built speculatively. They require commercial security through prior power purchase agreements (PPA’s) to offtake their generated electricity at a predetermined price. A pumped storage scheme with a large upper as at Lake Onslow is a logical provider of PPA’s because it requires significant power for pumping operation.

Unfortunately, coalition politicians and the gentailers will probably persist with their anti-Onslow renewable “chilling” argument right through to the 2026 election.

There would be reduced hydro spill.

The Onslow scheme operating in the electricity market would have the effect of the main hydro lakes being more often near their mid-ranges. This means that unpredictable sudden inflow events will more likely flow into hydro lakes that are not already near full, thus reducing the frequency of spill events at downstream power stations. This represents a positive national energy contribution from the scheme through more efficient operation of some hydro schemes.

Major hydro spills are not common in New Zealand, but they cause significant energy loss when they occur. A sequence of wet years can result in the Waitaki power scheme losing more energy from spill than the entire New Zealand hydro energy storage capacity. The Lake Pukaki spill rate at the end of December was equivalent to the power demand of Invercargill. Spills at our large dams like Benmore are impressive but represent renewable energy loss.

There would be reduced transmission loss.

Lake Onslow is located in the south of the South Island, far from the main electricity users in the North Island. However, the Onslow scheme is energy storage only and no extra power is sent north, so there is no additional transmission loss.

There would in fact be some reduction in transmission loss. A “dry year” is particularly with respect to the South Island, where most of the hydro storage and generation is located. At such times, the Onslow scheme would supply the shortfall of the South Island. That is, North Island power would remain in the North Island without the transmission loss of sending some if it south, as happened in August 2024 at the height of the dry winter (Fig. 3).

Figure 3: Electricity transfers between the North and South Islands in 2024. Southward transfer (black) dominates in August during the maximum of the 2024 dry winter. Vertical axis is in GWh. Source: Transpower.

The scheme would bring economic gains starting from its construction announcement.

The Lake Onslow scheme could take up to nine years to complete. However, major corporations work on long planning pipelines. The assurance of future relatively cheap and reliable green power is likely to result in electricity-sensitive industries setting up in Otago and Southland in anticipation of Onslow’s completion. New Zealand being in the process of building the world’s largest pumped storage scheme would also be a useful green flag to wave internationally.

An additional prior development derived from the scheme announcement would be new wind and solar generation being set up. This would be in anticipation of the new electricity market floor price, that would start from when the scheme became operational.

For extended dry periods, the scheme is a better economic alternative than demand reduction.
Demand reduction is when one or more major electricity users reduce their power demand by partial or total shutdown of operating plant when electricity supply is under pressure. This can be useful when the power supply issue is of short duration, such as when there is some major brief generation outage.

However, the use of demand reduction for more extended periods comes at an economic price. The resulting reduced product output may cause a corresponding drop in earning overseas funds. Also, extended load reductions have to be paid for, to the industries concerned.

For example, Meridian and Contact Energy would have had to pay some significant amount to Rio Tinto in compensation for the 330 GWh reduced power usage by the Tiwai aluminium over the period 10 June to 30 September, in the 2024 dry winter. If it had been operational, that 330 GWh could have been equally supplied by the Onslow scheme (storage capacity 5,000 GWh), and the smelter would have continued at normal production.

An additional advantage of the Onslow scheme over Tiwai load reduction is that, unlike the Tiwai smelter, Onslow’s power usage / generation could be quickly changed to reflect current power needs. In contrast, the smelter was not able to return to its previous production rate until April 2025, long after the 2024 dry winter.

An expanded Lake Onslow would be insurance against future extreme droughts in eastern Otago.
If the Onslow pumped storage scheme is constructed, its national justification would be to enable secure electricity at affordable cost for both homes and industry. However, there is also an East Otago regional water security aspect, making use of the large water volume at elevation as reserve against future extreme droughts.

The Dunedin city water supply, in particular, is vulnerable to an extended drought because most of the city’s water is derived from small inland catchments without significant storage and at risk of grassland fires. A narrow-diameter 20 km gravity-flow rock tunnel would link an expanded Lake Onslow to the Deep Creek Dunedin water intake. This could provide all of Dunedin’s water needs for 6 months without even an observable change in Onslow lake level. The emergency backup water in this case is analogous to Auckland having a reserve water supply via pumping through its water pipeline from the Waikato River.

On a larger water security scale, a short tunnel could link the expanded Lake Onslow with a headwater of the Taieri River. This is the main river of eastern Otago. It supplies both irrigation and town water along its length. In a drought emergency, the baseflow of the river could be maintained by releasing Lake Onslow water. Such a water release would need public consideration because of cultural and ecological issues from introducing Clutha water into the Taieri catchment. However, it would be logical to have the tunnel emplaced as part of Onslow scheme construction. This would give future generations the option of having emergency Onslow water available for the Taieri river if water supply became dire along its length.

Aside from the drought insurance aspect, the Onslow scheme would enable greater availability of some new irrigation water on a regular basis, both directly and indirectly.

In the direct context, the present Teviot Irrigation Scheme near Roxburgh would gain some additional water because the new Lake Onslow dam would eliminate the seasonal spill losses down the Teviot River that drains from the lake. That previously lost water could then be released at times to match irrigation need.

A larger Onslow water resource effect would arise in an indirect way. Specifically, the present seasonal storage operation of the Waitaki River hydro scheme has the effect of having increased the winter flow in the river, with some reduction at other times of year. The seasonal effect of the Onslow scheme in operation would be to take water away from the current winter flows, so that it becomes available to the river in summer for both irrigation, recreation, and ecological purposes.

The Lake Onslow pumped storage scheme would have local environmental and social impacts.

The Onslow scheme, if constructed, would extend the existing reservoir to flood some 60 square kilometres of Otago upland. Some farms would be impacted, and areas of natural wetland would be lost.

During the construction phase, the region around the new Onslow dam would be an industrial zone, removing the solitude valued by the local people for fishing and water recreation.

Also, the scheme gains its energy storage capacity by having a large operating range. At times of maximum drawdown, the new lake would still be larger than at present but surrounded by a wide zone of rocky dried former lakebed. Dust issues could be a possibility unless there is prior removal of the thin soil on the schist rock within the operating range.

Farms cannot be easily compensated in that it would be difficult to find an equivalent farm on the market for those affected.

Some other impact aspects could have offsets of various types. For example, the people of Roxburgh and the surrounding area might reasonably request a new road be constructed to give direct access to the nearby Upper Manorburn reservoir, which has a remote setting similar to Lake Onslow.

Also, it may be possible to create new wetland reserves in the upper Taieri River as compensation for the lost Onslow wetlands.

One positive environmental consequence of the Onslow scheme (other than emissions reduction) would be a partial return of the lower Waitaki River toward its pre-hydro state of higher spring and summer flows. This would arise from releases of water from Lakes Tekapo and Pukaki to provide seasonal power for pumping.

Other aspects of the scheme would be regarded with mixed views. Some will welcome the creation of new employment opportunities, but others would see it as lifestyle disruption.

Figure 4: the lower Waitaki River has been impacted by upstream hydro operations.

Selected references

Reports produced from MBIE NZ Battery investigations are currently listed under the heading “Lake Onslow pumped hydro”.

Work on the Lake Onslow pumped storage scheme was initiated at the University of Waikato. No external funding was involved. Relevant references are listed below.

Karaminik, Y. (2024). Environmental offsets and water resource opportunities with Lake Onslow pumped storage (MSc).

Bardsley, W.E., Karaminik, Y., Majeed, M. (2022). Estimating Teviot River compensation flow to offset evaporation loss from Lake Onslow pumped storage. Journal of Hydrology (NZ), 61, p.179-182.

Kamarinik, Y., Bardsley, W. E. (2020). Onslow pumped hydro: environmental offsets and spinoffs. New Zealand Hydrological Society Annual Conference, Invercargill.

Majeed, M. (2019). Evaluating the potential for a multi-use seasonal pumped storage scheme in New Zealand’s South Island (PhD).

Bardsley, W.E., Majeed, M. (2015). A multi-functional large pumped storage scheme for New Zealand in support of renewable energy development? International Congress on Modelling and Simulation (MODSIM2015). Gold Coast.

Majeed, M., Bardsley, W. E. (2015). Assessment of economic and environmental advantages of a seasonal pumped storage scheme (Onslow, Central Otago). New Zealand Hydrological Society Annual Conference, Hamilton.

Majeed, M., Bardsley, W. E. (2014). Simulation models to evaluate economic feasibility of a possible seasonal pumped storage scheme at Onslow, Central Otago. Electricity Engineers Association Conference Wellington.

Majeed, K. M., Bardsley, W. E. (2012). Prospects for pumped storage in Central Otago. New Zealand Hydrological Society Annual Conference, Nelson.

Bardsley, W. E. (2006). A pumped storage scheme for maintaining hydro electricity against climatic variations? (Invited paper) 8th Annual New Zealand Energy Summit, Wellington.

Bardsley, W. E. (2006). China and New Zealand: Large seasonal pumped storage schemes for an electricity future using wind power and small hydro systems? International East Asia Regional Workshop of the International Academy Panel (IAP) on Water Security with Climate Change and Human Activity, Beijing.

Bardsley, W. E., Leyland, B., Bear, S. F. (2006). A large pumped storage scheme for seasonal reliability of national power supply? Electricity Engineers Association Conference, Auckland.

Bardsley, W. E. (2006). The Onslow seasonal pumped storage scheme revisited: An alternative approach to water level management of the southern hydro lakes? New Zealand Hydrological Society Annual Conference, Christchurch.

Bardsley, W. E. (2005). Note on the pumped storage potential of the Onslow-Manorburn depression, New Zealand. Journal of Hydrology (NZ), 44, p.131-135.

Bear, S. (2005). Hydrological evaluation of pumped storage in the Onslow-Manorburn basin (MSc).

Bear, S. F., Bardsley, W. E. (2004). A large New Zealand pumped storage scheme for reliable power through dry years? New Zealand Hydrological Society Annual Conference, Queenstown.

Bear, S. F., Bardsley, W. E. (2003). A pumped storage/thermal station hybrid for maintaining New Zealand electricity supply through dry years. New Zealand Hydrological Society Annual Conference, Taupo.