At present, even one period of shortage creates political and public outrage, so the chance of scarcity prices being seen is tiny:  in effect, the costs to market participants of actual shortage is already substantially higher than scarcity prices, even at their current levels.

But this illustrates one of the key challenges of getting the market to 100%RE:  maintaining a secure supply while also having prices that are efficient in terms of signalling the need for new generation.  The thermal buffer ensures that when demand grows there is not only spare capacity available to keep the lights on, but a more measured increase in prices (as opposed to going straight to scarcity prices).  But if nothing changes, when the thermal buffer is gone, things will be very different.

A corollary of this is that when we think about how the market needs to change to get it to 100%RE, we need to focus on what new mechanisms might be required to set prices at levels that will incentivise new generation as it is required.

A second corollary is that there should be an increasing focus on what assets, contracts and mechanisms will replace the current thermal buffer, which could include biofueled-thermal generation, demand response, storage, renewable or zero-emission hydrogen, flexible geothermal, nuclear, and options that may not have been thought of yet.

Firming and peaking

Returning to the longer story, by firming, I mean the small fleet of fossil-fueled generators that start up and run for extended periods, sometimes several months, when inflows into the hydro lakes fall and hydro lake levels follow, i.e. during dry spells.

By peaking, I mean these same fossil-fueled generators along with potential alternatives including battery energy storage systems (BESS) and demand response, the latter being when consumers turn down their demand during short periods of market stress.  Peaking generators (or demand that can be reduced) are fast-acting when called on during brief periods of high demand, such as the morning and evening demand peaks, hence the name.

The firming sector is highly concentrated in the four large, vertically integrated gentailers (generator-retailers), and only remains in the market because these companies have large portfolios of generation and customers, and the financial strength to pay the fixed costs of maintaining the assets making up the thermal buffer, even when they may not get used for long periods.

As noted above, the fixed costs associated with keeping generation available to the market are not covered by explicit additional payments for capacity, at least not payments that are defined in the rules of the market (referred to as the Electricity Industry Participation Code, or just the Code).

But Channel Infrastructure at Marsden Pt, is considering diesel-fired peaking generation at its site as an alternative to importing liquified natural gas (LNG). Channel was recently quoted on the Energy News website as saying a peaker would require a “capacity-based payment structure for it to work” which refers to a regular payment to keep the peakers available, whether they are used to generate or not.

Since around 2010, the gentailers have had various “swaption” arrangements in place, starting with a swaption between Meridian Energy and Genesis Energy.  A swaption is an option to buy an industry-standard swap contract, often referred to as a contract-for-differences or CFD which, in return for a fixed annual payment known as the option fee (or premium), gives the buyer of the swaption the right to “call” the CFD during the periods, and at the prices and quantities specified in the swaption contract.  The seller of the swaption was Genesis, and when the swaption was called it would start up a fossil-fueled generating unit, often one of the steam turbine (a.k.a. Rankine) units at Huntly.  The option fee was paid to help to keep the Rankine units in service, and so was a form of capacity payment.

Genesis Energy has offered two forms of option contracts (market Security Options in 2022 and Huntly Firming Options in 2024) which would allow buyers to (notionally) call on generation in return for payment of a fixed annual premium:  again, the premium is a form of capacity payment.

NZ Aluminium Smelters (NZAS) signed new contracts with Meridian, Contact and Mercury on 31^st May 2024 to keep the Tiwai Pt smelter running, and two of the contracts (Meridian and Contact) provide a mechanism which is analogous to a swaption backed by generation, in which the hedge commitments of these gentailers are reduced when called in return for, you guessed it, a fixed annual premium, another form of capacity payment.

These capacity payment examples show that, despite the fact that the NZ market is supposedly energy-only, a shadow capacity market already exists.  It has existed for over a decade, and it is growing.  Some of this would see the thermal buffer remain in the market, or potentially expand, but some of it  now relates to demand response.

But the shadow capacity market operates, as the name suggests, more or less out of sight, (technically, the capacity market operates in the wider electricity hedge market) and the dollar amount of the premiums are generally not disclosed.  What we do in the shadows tends to remain in the shadows and hence does not create transparent price signals which could, for example, facilitate entry of new players into the peaking and firming sector with offerings with lower premiums.

The thermal buffer is ultimately at the mercy of decisions made by the gentailers for their respective business purposes, which may or may not lead to an optimal buffer remaining in the market for the long-term.

The traditional thermal buffer will shrink as we move closer to 100%RE, which means its energy revenues will shrink, making it ever more reliant on capacity payments, currently obtained in the shadow capacity market.  Given the criticality of the buffer to both security of supply and pricing, it is time to decide if the market needs to abandon its pure energy-only model and allow for capacity payments to be made to the owners of peaking and firming plant, and for the total of all these capacity payments to be recovered from the market in some way.

In principle, a capacity market would operate by setting a total MW capacity requirement and then auctioning the rights to receive capacity payments.  Intermittent renewable generation could bid for a small percentage of their respective capacity, but fossil-fueled thermal generation could bid for most if not all its capacity, since this would be considered firm capacity.

Who would set the MW requirements?  Probably the regulator, which immediately introduces the possibility of over-investment in capacity, because the regulator is (a) unlikely to be any better at forecasting demand (including capacity requirements) than anyone else and (b) would have an incentive to over-contract to avoid public and political backlash if capacity ever proved to be lacking.

Allowing all generators to participate in a capacity market introduces the vexed issue of how much wind and solar contribute.  In terms of peaking, relatively little.  But in terms of firming, their contribution would be much greater as dry spells typically last weeks or months, during which time even intermittent renewables exhibit relatively stable energy production profiles.

Furthermore, it is all very well to contract capacity but then how would the market know that the capacity was actually available when required?  And what if not available when required?  These concerns suggest that a new and strong compliance monitoring task would be required of the regulator, with financial penalties applied to parties that fail to deliver capacity when required.

A capacity market does not have to cover all generators, however.  In principle, it could just cover those that (a) provide fully reliable peaking or firming, or both and (b) are in danger of exiting the market if their fixed costs are not covered by some alternative means.  This would limit the number of capacity market participants and therefore reduce the administrative and compliance-related burden.  It would need to focus on the capacity that is required to meet peaks, after netting expected renewable generation off peak demand, and also the energy that is required during dry spells when inflows into the hydro lakes sit below average for months at a time.

Rather than have the regulator determine how much peaking and firming is required, it would be desirable to have the market determine how much capacity should be contracted, although the regulator could have a role in providing information relevant to this process.  Most likely, responsibility for ensuring they have cover during peaks and dry spells would rest with the electricity retailers.

For example, at present electricity retailers are not required to have hedge cover for peaks and dry spells, but the Code could be changed to require all electricity retailers to contract this hedge cover, which would have the effect of the market collectively contracting sufficient capacity to cover peaks and dry spells, based on the market’s aggregate assessment of future demand.  There is currently a requirement for retailers to report quarterly on their exposure to extreme scenarios of scarcity prices and dry spells (the relevant rules in the Code form what is known as the “stress testing” regime), which could be strengthened to require retailers to demonstrate a minimum level of hedge cover in these scenarios.  To meet the minimum requirements, retailers would have to demonstrate they have cover through a mix of hedges backed by thermal generation, demand response, access to storage energy, and any other relevant measures that develop over time.

But there are likely many ways this could be implemented, and I don’t have all the answers as to how such a mechanism would operate in the NZ context.  Indeed, that would need to be worked out carefully over a period of market redesign including extensive input from the market and the public.

But I do believe transparency around the shadow capacity market needs to be improved as soon as possible, so the costs of maintaining peaking and firming capacity in the market can be seen by any parties that might look to enter the shadow capacity market with peaking or firming options of their own.  Disclosures would need to extend to capacity payments, fuel contract details, and other contract details including the level of contracting so that, for example, plant that is uncontracted is not retired with little or no warning.

Improving the arrangements around firming would give the market and the regulator a greater degree of oversight and potentially some control over how much firming capacity and fuel is available, thus improving security of supply.  For example, if peaking and firming capacity is contracted to a low level, and in danger of being retired, the regulator could implement measures to ensure costs are covered until such time as contracts are in place, or mandate contracting to minimum levels.

Scarcity pricing improvements

There is a problem with scarcity pricing.  When sudden and unexpected shortages do occur, which is very seldom, then there is huge public and political push-back.  As a result, the market does everything it can to prevent such occurrences.  As a result, retailers and spot-exposed consumers don’t pay these prices and generators don’t receive them.  As I noted above, the wider costs of shortage are probably already much higher than scarcity prices, so scarcity prices are more-or-less ignored.

So, what is their use?

Well, the answer must surely be that they serve no useful purpose at all in their current form, unless the threat of scarcity prices being struck in the market causes market participants to change their behaviour over and above other measures taken to avoid shortage:  I doubt this is the case.

Which then begs the question of whether they should either be scrapped or revised so that they do serve a useful purpose. 

One further problem with scarcity prices is that they can be a massive step change from market-based prices when they do occur.  Offer prices currently in the market max-out at $6,650/MWh which is only 32% of the lowest scarcity price of $21,000/MWh.  The gap can be even larger if a constraint on the grid causes a region of the grid to become effectively isolated from the rest of the market, e.g. a constraint occurs, creating a shortage of generation in the region downstream of the constraint, but the only generation downstream of the constraint is priced at $0.01/MWh, e.g. a windfarm or solar farm.  If a shortage results, there would be a difference of at least $21,000/MWh.

To have any effect when scarcity prices are virtually never struck, the scarcity pricing mechanism needs to change so they apply in advance of a shortage, i.e. in response to the increasing probability of a shortage occurring.  For example, if we consider the “supply margin” to be the gap between total generation offers in a period and forecast demand, then calculated spot prices would be increased in response to a shrinking supply margin by an amount proportional to the gap between the highest relevant offer price and $21,000/MWh:  let’s call this “progressive scarcity pricing”.  This adjustment might be made, for example, once the forecast supply margin falls below 100 MW, just to pick a number out of thin air. The adjustment could be linear, or it could be relatively small initially then increase sharply as the forecast supply margin approaches zero.

The two examples shown in the chart below have the supply margin (total generation offered less forecast demand) fall to 26 MW in trading period 35, and 87 MW in period 37.  In the chart on the left the margin goes to -27 MW in period 36 (actual shortage) and on the right it goes to 5 MW (not quite shortage).  The blue columns are the prices set under the current market rules, and the orange columns are ‘adders’ worked out on an exponential basis, i.e. the adder is small when the margin is close to 100 MW but increases exponentially as the margin approaches zero.

In the example on the left, the price spikes in period 35 but then hits $21,000/MWh when actual shortage occurs.

In the example on the right, the current market settings would see the price remain at $1,000 in period 36 despite being right on the edge of shortage, i.e. the current scarcity pricing mechanism is grossly binary, being not there one period and then there the next[3].  But using progressive scarcity pricing, the chart shows the spot price spiking toward the lower bound of scarcity prices, indicating a high probability of shortage.

A mechanism such as this would better reflect how close the market is to actual shortage and put scarcity prices into the market on a more regular basis before an actual shortage occurs, contributing to the signals to bring more generation to the market, thereby improving security of supply. 

Furthermore, an enhanced scarcity pricing mechanism would contribute to the price signals without actual shortages occurring.  Most importantly, an enhanced scarcity pricing mechanism would help to signal the need for new generation to be built as the thermal buffer shrinks.  As the market moves closer to shortage, scarcity pricing would start to raise spot prices ahead of actual shortage.  Ultimately, this could help to avoid the nightmare scenario of an equilibrium developing in which shortage becomes a persistent feature of the market, due to the need to have prices which incentivise new generation as it is required.

Demand-side response

Before I move onto the next major issue, I need to briefly mention our proposal for “balancing demand” to be added as a new type of generation, described in an article on our web site called Unleashing Demand Response, and submitted in April to the Electricity Authority as a Code amendment proposal. 

This would allow qualifying demand to be offered as generation, providing the opportunity for larger consumers to contribute to keeping the lights on in a way which rewards the consumers for making the demand available, including compensation for their costs in doing so.

Implementing balancing demand would provide another mechanism which could set prices that would signal the need for new generation without having physical shortages.

Taking a wider view, balancing demand should be seen as one way of rewarding consumers for providing the equivalent of generating capacity during peaks or dry spells, or both.  A key feature of the balancing demand proposal is the requirement to offer as generation, which dispatches the balancing demand at its offer price or higher.

Balancing demand is targeted to short-term demand response, e.g. meeting peak demand, as being dispatched on and off for months during a dry spell is potentially highly disruptive to a consumer’s business.  Demand that is turned off during dry spells would help to replace the thermal buffer, but would need to receive a capacity payment in return for being available.  This would not fit with our energy-only market, as a capacity payment cannot be dispatched and cannot set spot prices. 

Paradoxically, reducing demand without being dispatched depresses spot prices, because the dispatch model forecasts lower demand, which in turn leads to less generation being dispatched and lower prices.  This is the opposite of what is desirable when more generation is needed to meet demand during dry spells, i.e. the pricing signal from dry-year demand response should ideally be reflected in spot prices, but while it operates in the shadow capacity market this pricing signal is lost.  This would be difficult to achieve with longer-term demand response, which operates in the shadow capacity market, but would benefit from greater transparency suggested above.

Customer Compensation Scheme (CCS)

The CCS was introduced in 2018 and is defined in the Code. It requires electricity retailers to pay qualifying customers $12 per week when they are asked to reduce electricity consumption during an official conservation campaign (OCC), which would typically happen during a prolonged, severe dry spell.  For example, a small independent retailer with 50,000 residential customers would pay its customers $600,000 per week in aggregate.

Since its introduction, hydro storage has not gotten low enough for an OCC to be called, so the CCS rules are as-yet untested.  This may be because the CCS also applies to the gentailers that operate the storage and the thermal buffer, and it is in their financial interests to avoid this payment. 

I have never understood why the CCS applies to all retailers, because small independent retailers have no generation and no control, direct or otherwise, over hydro storage.  So, if they ever have to pay customers during an OCC, it will be through no fault of their own.

On the other hand, the four large gentailers do have storage and thermal generation, and so storage is under their collective control.

Ideally, the CCS would only apply to large gentailers, but if an OCC ever did arise, customers of other retailers would not be compensated for their voluntary savings, which hardly seems fair.  Therefore, if the CCS is retained, it should probably has to apply to all retailers, but should also be part of an enhanced stress testing regime described above, i.e. the stress tests should include one or more tests in which an OCC is called.

New Generation

I often hear in the media that the market needs to build more generation to keep the lights on and to keep prices down, but I feel these statements overlook what is actually happening.

The chart below shows the annual energy output of all new renewable generation built from 2009 to 2025[4], and it shows the rate of build has increased significantly since 2020/21, with the annual rate of increase particularly high in 2024 when 1,472 GWh of new generation operated (the next highest was 1,362 GWh in 2010).  New gas-fired thermal generation was also built (in 2010, 2012 and 2020) but is excluded from the chart because its aggregated output changes in response to wet and dry spells, whereas wind, geothermal and solar do not:  if this were added in, and wet and dry spells were corrected for, it would show the same trend.

To be absolutely sure the rate of build has picked up, we also need to consider the margin between total supply and demand, because if demand is growing faster than new generation is built, the picture is not so positive on the supply-side of the market.  The next chart shows the total GWh per annum built, now with and without new thermal generation included, along with the total metred demand reported each year by MBIE[5].  Demand growth was watered down in 2024 by demand reductions at the Tiwai Pt smelter in response to dry spells, so these demand reductions are added back in.  The chart shows that the energy supply increased throughout the period while demand remained relatively constant, albeit with significant year-to-year fluctuations.

So, the evidence suggests that new generation is arriving in time to meet new demand growth, whenever that eventuates, at least when measured on an annual basis.  In 2024 there was a large and unexpected drop in gas supply, which coincided with a prolonged and deep dry spell during which hydro generation was lower than average, to conserve storage, combined with a relatively calm period in respect of windfarms:  these events masked the fact that renewable generation is increasing rapidly.

Gas-fired Electricity Generation

On 5^th June, MBIE released the annual natural gas reserves data, as they were assessed on 1^st January 2025 and the chart shows the reserves back to 2000 in PJ[6].  The fall since 2022 is 1,050 PJ (53%) and if this fall continues then gas supplies could be exhausted by late 2027.  That rate of decline is not particularly likely, for technical and other reasons, but the reality is that NZ is very short of gas.  The last few years have shown that gas can be made available for electricity generation, at the expense of methanol production at Methanex’s Motonui plant in Taranaki. 

Electricity is a premium energy source, and its high value ensures that electricity generators can out-bid other large gas users, Methanex and other industrials, when push comes to shove.  But continuing to do so will cause electricity prices to remain higher than otherwise, due to gas’ influence on electricity (covered in the next section on Electricity Prices), and will put financial pressure on all other gas consumers to the point where industrial consumers may have to cease operations in NZ, and smaller commercial and residential consumers will end up paying higher prices.

Finding more gas is an option, but 2027 appears to be the earliest any significant new field[7] can be brought into production, and larger fields could take a decade or more to develop, assuming they prove to be commercially viable after more drilling is done.

Gas-fired electricity generators acknowledge that their reliance on gas was misplaced, and given they have alternatives available to them in the form of new renewable generation, the only viable strategy for them is to move away from using large amounts of gas for generation as quickly as possible, which also leaves more gas for other users. 

For example, Genesis Energy is working to source large quantities of wood pellets, a renewable fuel, to burn in the Rankine units at the Huntly power station, and to retain the ability to burn coal and gas at the Huntly power station when required.

Given the dire situation the gas market finds itself in, one role I see for government is to ensure that the move away from gas is made as quickly as possible, potentially by increasing the rate at which renewable generation is built, even beyond where it is now.  We’ve already seen the renewable build is increasing anyway, so intervention may not be necessary, but if there is a role for government at all, it could be to provide incentives to all existing and intending new generators to commit to building as soon as possible.  There’s a lot of detail here to work through, obviously, including whether intervention is needed at all, how much would be needed and for how long, and what form would it take.  So, I am effectively flagging this as a key issue to watch.

Electricity Prices

There is currently much talk of how the market being dominated by four large gentailers reduces competition.  But even with more competition, two factors could act to keep prices relatively high for the medium term.  One is the price of gas, and the other is the cost of building, owning and operating new generation.

The prevalence of gas-fired generation in the thermal buffer means that electricity prices are driven to a large extent by gas prices, as shown in the first chart below, which shows annual average electricity contract prices for larger consumers and wholesale gas prices going back to 2009.  There are good reasons for this linkage:  gas generators can be the marginal producer when they run, and water values are heavily influenced by the cost of generating at gas-fired thermal power stations.  The correlation is almost perfect (0.97) so we have both causation and correlation.

As gas supplies were interrupted or fell after 2018, prices rose, and concerns developed over the ability of the electricity market to keep the lights on, while consumers experienced, or will experience, upward pressure on prices.

Is it likely the price of gas will come down significantly any time soon?  No.  NZ is no longer seen as an attractive place to invest in gas exploration, and more exploration is required because fields that are already producing are mostly in steep decline, and hence the gas supply crisis of 2024.

Even if more gas is discovered, gas wells operate optimally when draw-off is relatively steady.  But the thermal buffer needs to be increasingly flexible as it shrinks, so electricity generators need to invest in more gas storage so that gas can be stored when not required, and then taken out of storage when it is required. All this adds to the cost of using gas to generate electricity.

If finding and using more gas is increasingly difficult and expensive, the obvious solution is to build more wind, solar and geothermal, thus reducing reliance on gas.  Adding more supply naturally tends to lower prices, if all other things remain equal.

The next chart adds the estimated real LCOEs (levelised cost of electricity), in 2025 dollars, of each of the new plant that was actually built in New Zealand since 2009, or is under construction, using data in the public domain.  The LCOE of a new generation project is the constant average price attained for a generator’s output over its lifetime that would just cover all cash costs and provide a fair return to the generator’s owners.

We observe a falling trend in LCOEs through to 2020/21 but since then LCOEs have risen sharply and are now between $110/MWh and $153/MWh based on the data points from 2024 and 2025.  This is a wide range, making it difficult to put a firm number on the actual increase in LCOEs, but just using the annual averages (excluding 2025 which only has one data point) the increase is 45% in real terms or 69% in nominal terms.

It is critically important to use actual projects in New Zealand rather than relying on LCOE curves which may be relevant to other nations but not to NZ, for whatever reason.

The charts illustrate how the combination of rising cost of gas and rising cost of new generation have led to electricity price rises.

Notwithstanding increases in LCOEs, the New Generation section showed a significant increase in the amount of new generation built in 2024, which is higher than Energy Link forecasts in our quarterly Price Path.  In the April 2025 Price Path we included an “Outside the Box” (OTB) scenario which projected the current of rate of build through to 2031, and reran this in the Price Path Base Case.  The chart below shows the additional build in GWh per annum and the difference in price between the Base Case and the OTB, which increases progressively through to about $12/MWh (lower) in 2031, i.e. the market is currently building new generation at a higher rate than we forecast for the next few years.

But just because the rate of build is currently higher than forecast does not mean it will continue this way into the future, but the discussion in the Gas-fired Electricity Generation section suggests the dire state of the gas supply situation could cause the build-rate to remain higher than otherwise expected, for some time yet.  It is also conceivable that government may intervene to ensure this happens, freeing up gas for other users.  We won’t know until the government makes decisions, if any, once they have considered the review by Frontier Economics along with other advice and presumably a period of public consultation.

But returning to the issue of how to make electricity more affordable, there are a limited number of options:

1.              increase the supply of gas until gas prices fall;

2.              reduce reliance on gas:  as gas-fired generation falls, gas contributes less to      the average spot price;

3.              find cheaper alternatives to replace today’s thermal buffer;

4.              build more renewables than would be economic based on spot price             expectations;

5.              reduce LCOEs and the cost of connecting to the grid;

6.              encourage greater competition in building new generation.

Significantly increasing the gas supply will take years, so rapidly reducing reliance on gas is the only strategy that makes sense.

Genesis Energy is working hard to develop the capability to run its Rankine units on wood pellets, which will be produced in NZ.  So far, the company has said the cost of pellets is likely to be similar to the cost of burning coal and paying the associated carbon costs, but exactly where the price level will fall is still uncertain.

But in this section I will focus on LCOEs and whether or not new generation could feasibly be built at a rate greater than it might otherwise.

As I outlined at the start of this article, developers rely on prices that will make their projects viable, so they must wait until expectations of prices rise enough to meet this test.  That is perfectly rational behaviour in any market.  Any business wishing to establish new products or to otherwise expand will wait until it is confident that there is growth in demand that will create the market for its expanded offerings, before making the necessary investment and procuring the additional human and other resources required to expand.

It may well be that changing market settings to create more competition for the gentailers will lead to more generation being built by developers that are independent of the gentailers.  Competition increases the incentive to reduce costs and to innovate, finding new ways to solve existing problems.  All of this could lead to generation being built at a higher rate, although it does not provide incentives to build at a rate which depresses prices below levels that will support that new generation.

One alternative market structure proposed in the past was that of a single buyer for generation, in which the regulator must forecast when demand will grow, and by how much, and then seek proposals from new generators to build generation to meet that demand growth.  Successful bidders are then paid an agreed annual price for their energy, for at least several years, to guarantee a return.

There are all sorts of problems with this model, including the inability of the regulator to forecast demand accurately, the incentive of the regulator to over-build, the lack of competition during a generator’s fixed pricing period, and bias on the part of the regulator to pick projects conservatively, which could stifle innovation by trying to ‘pick winners’.

What has not received a lot of attention to date is the large rise in the LCOEs of new generation. This is a key issue, because if ways can be found to reduce the cost of generation, then prices should come down even if there is no change to the current market settings or structure.

We already know that the cost of equipment increased after covid, as has the cost of construction, i.e. those costs not directly related to plant and equipment.  But I believe there needs to be work done by the Electricity Authority to determine where costs have increased, by how much, and whether suppliers to new generation projects operate in a competitive market in NZ. 

For example, it may be that suppliers of a particular generating technology operate with limited competition, which could make it difficult for other suppliers to gain a foothold here.  It’s easy to see how a monopoly could develop in our small market:  a supplier gains a track record of several projects and establishes the reputation of its equipment for reliability and availability.  Developers tend to be conservative because their projects put hundreds of millions of dollars at risk, so once a supplier has this reputation other suppliers find it hard to convince developers to use their equipment or other services.

Looking at the windfarm project LCOEs included as data points in the chart above, only two wind turbine suppliers, Siemens and Vestas, are represented, with Meridian Energy favouring Siemens and Mercury favouring Vestas.  The lack of a third, fourth or fifth supplier does not necessarily mean the market for turbines is not as competitive as it could be, but the question should be asked.

In addition to establishing why costs have risen and whether the supplier market is competitive, it may be time to consider whether any interventions by government would be better targeted at helping to reduce LCOEs instead of or in addition to modifying the behaviour of electricity market participants.  I am no expert in this area, but I can imagine that funding could be made available to developers prepared to try a new supplier, for example.  Perhaps NZ needs a Pharmac of generation technology that could establish better terms for key pieces of equipment?

Non-equipment and non-plant costs also need scrutiny.  The media frequently covers the costs and delays of consenting projects, but once consented I wonder if NZ has the required capability and capacity available to install various types of new renewable generation at scale, at a reasonable cost?

Finally, the cost of capital (debt and equity) is a major factor in determining LCOEs, so it would be useful to better understand if gentailers have lower or higher costs of capital than other developers, how easy finance is to access, and what might help to lower the cost of capital for new generation projects.

One thing is for sure, unless efforts are made to understand why LCOEs have increased, the regulator, government and market participants won’t have the data to make decisions on whether there are measures which would help to reduce them.

Summary

Putting this all together, I believe the following package of measures should be implemented to give the best chance of managing and eventually replacing the thermal buffer as the market moves closer to 100%RE:

1.           shadow capacity market:  ensuring greater transparency around capacity-based payments and mechanisms relating to demand peaks and dry spells;

2.           contracting: investigating a requirement for all electricity retailers to have contracts in place with capacity providers;

3.           reducing gas: moving away from reliance on gas-fired generation as quickly as possible;

4.           scarcity pricing: changing the scarcity pricing regime to better signal the rising probability of shortage;

5.           CCS: reviewing the Customer Compensated Scheme to determine if it should be retained;

6.           demand response:  rewarding consumers for providing demand reductions during peaks and dry spells, including through a mechanism such as balancing demand.

Having a mix of balancing demand and progressive scarcity pricing, along with batteries, would go a long way to replacing the peaking functions of the existing thermal buffer.

Replacing the gas-fired firming generators would require additional demand response which receives capacity payments in return for being removed from the market for extended periods.

In respect of electricity prices, any measures that lead to an over-build of renewables (relative to what the market is already delivering), including a capacity market or single-buyer model, will increase the overall cost of supply, while simultaneously reducing prices in the spot market.  Costs not recovered through spot revenue will need to be recovered through alternative charges:  ultimately, consumers and politicians will have to decide if any increase in total cost is justified by an increase in the security of supply obtained by an over-build, relative to what the market is already delivering.

Encouraging greater competition in building new generation could help to bring prices down by putting greater pressure on developers to reduce costs, and spurring innovation across the supply chain.

In addition, substantial effort should be made to understand why LCOEs have increased so dramatically since 2020/21, and hence where there are bottlenecks, or insufficient competition, or a lack of capable resource in NZ.  Such an undertaking would provide the data necessary to evaluate, design and assess measures which could help to bring LCOEs down, which would ultimately lead to lower electricity prices, and simultaneously either increasing security of supply.  This effort would require detailed information gathering by the Electricity Authority, followed by a period of analysis by the Authority, potentially with assistance from the Commerce Commission in respect of competition between suppliers to generation developers.

New Zealand is in a unique position within the global transition to renewable electricity, with large hydro, geothermal and wind resources already well developed, with solar catching up fast, but lacking connections to a wider continental electrical grid, as in Europe or North America. As we have seen, this presents unique challenges as we move away from fossil-fuelled generation which historically provided both energy security and pricing stability. It is only appropriate that as these significant changes are made to the physical supply system, that the market adapts correspondingly.

---

[1] The Maui East field offshore Taranaki.

---

[2] 1 PJ is approximately equal to 278 GWh.

---

[3] There is not enough data for 2025 to make a reliable estimate.