Methodology

What goes in

Beacon uses public New Zealand electricity-market, transmission-network and geospatial datasets as source inputs to its screening model. These datasets are used as source inputs only. Lumere independently compiles, cleans, joins, calibrates, weights and interprets the data to produce Beacon's scores, classifications, rankings and commentary.

• 1. Transpower interconnection branch-capacity data. Beacon uses published Transpower interconnection branch-capacity files to identify indicative MVA ratings on transmission corridors. These files are published under the Electricity Industry Participation Code 2010 and are reusable under the Creative Commons Attribution 4.0 International Licence, with attribution to Transpower New Zealand Limited, as confirmed by Transpower in August 2026. Beacon uses this information as an indicative asset-capability input only. It does not represent confirmed transfer capacity, available capacity, operational grid capability, or project-specific connection capacity. Updated monthly where new files are published by Transpower.
• 2. Electricity Authority EMI nodal pricing and reserves. Beacon uses three years of half-hourly historical settlement prices at grid exit points to derive nodal price, spread, volatility and historical market-signal indicators, and three years of half-hourly reserve prices to derive island-level fast and sustained instantaneous reserve indicators. This information is used to identify historical market signals only. It does not forecast future prices, revenues, dispatch outcomes, or investment performance. Updated monthly.
• 3. Electricity Authority EMI generation and plant data. Beacon uses published metered generation output, embedded generation and dispatched generation plant data to identify existing generation at each connection point, and to measure what generating plant has historically earned relative to the average price. These measured indicators are Lumere-derived and are calculated only where sufficient metered history exists; where it does not, Beacon says so rather than substituting a national figure. This information describes historical performance only. It does not forecast future generation, capture, dispatch outcomes or revenue. Updated monthly.
• 4. Transpower Envision Opportunities information. Beacon uses publicly available Transpower Envision Opportunities information as a high-level input to its substation headroom screen. Beacon may derive or classify N-1 or headroom indicators from this information where available. These indicators are Lumere-derived screening outputs. They are not official Transpower outputs, not confirmed or available capacity, not Transpower advice, and not a substitute for a project-specific feasibility study or Transpower's connection process. Transpower New Zealand Limited has not reviewed, approved, endorsed or validated Beacon, Lumere's methodology, or any Beacon ranking, classification, score or recommendation. Any connection remains subject to Transpower's connection process and detailed technical assessment. Users should engage directly with Transpower and obtain appropriate technical, commercial and legal advice before making connection, development or investment decisions. Updated annually where new information is published by Transpower.
• 5. Transpower generation and battery connection pipeline. Beacon uses published Transpower generation and battery connection pipeline information, including announced and anonymised projects, project stage and MW where available. This information is used as a contextual and screening input to assess future connection activity, queue pressure, potential competition for capacity and development context. Connection pipeline data is reusable under the Creative Commons Attribution 4.0 International Licence, with attribution to Transpower New Zealand Limited, as confirmed by Transpower in August 2026. Beacon reproduces and adapts selected connection pipeline information as part of Lumere's independent methodology. The connection pipeline is indicative, subject to change without notice, and does not represent committed, approved, reserved or guaranteed connections. It does not indicate the availability of grid capacity at any location, and should not be relied on as evidence of the existence, progress, timing or likelihood of any connection or third-party project. Updated monthly where new files are published by Transpower.
• 6. Transpower load, storage and other non-generation connection pipeline. Beacon uses published Transpower load, storage and other non-generation connection pipeline information as contextual information on future demand, electrification, storage activity and local network pressure. Connection pipeline data is reusable under the Creative Commons Attribution 4.0 International Licence, with attribution to Transpower New Zealand Limited, as confirmed by Transpower in August 2026. Beacon reproduces and adapts selected connection pipeline information as part of Lumere's independent methodology. The connection pipeline is indicative, subject to change without notice, and does not represent committed, approved, reserved or guaranteed connections. It does not indicate the availability of grid capacity at any location, and should not be relied on as evidence of the existence, progress, timing or likelihood of any connection or third-party project. Updated monthly where new files are published by Transpower.
• 7. Transpower public GIS. Beacon uses Transpower public geospatial data for substations and transmission-line geometry, including mapping, spatial joins and corridor association. Transpower's open geospatial portal states that most of its public geospatial data is available under the Creative Commons Attribution 4.0 International Licence and must be attributed on maps, websites, applications or printed material using the data. Beacon uses Transpower public geospatial data with attribution to Transpower New Zealand Limited and in accordance with the licence or terms stated for the relevant geospatial dataset.
• 8. Stats NZ regional boundaries. Beacon uses Stats NZ regional boundary datasets for mapping, spatial analysis and regional grouping, including Regional Council 2025 boundaries where relevant. Stats NZ's Regional Council 2025 dataset is listed as Creative Commons Attribution 4.0 International and permits commercial use with attribution.

Beacon's forward views are produced by Lumere's own hydrology-normalised fundamental model. They are not a third-party price curve and are not a forecast.

Sources and licensing

Beacon is built from publicly available datasets published by New Zealand electricity-sector and government data providers. Lumere uses these datasets as source inputs only. Beacon's compilation, cleaning, joining, calibration, scoring, classification, ranking, commentary and commercial interpretation are Lumere's own work.

Where a source dataset is expressly licensed under the Creative Commons Attribution 4.0 International Licence, Lumere uses it under that licence and attributes the relevant source. Where Transpower has confirmed that a source dataset is reusable under the Creative Commons Attribution 4.0 International Licence, Lumere uses that dataset under that licence, attributes Transpower New Zealand Limited, identifies where Lumere has adapted or transformed the material, and does not imply endorsement by Transpower. Where a dataset is published without an expressly stated reuse licence, Lumere attributes the source and treats use as subject to any terms stated by the relevant publisher. Lumere welcomes contact from any data publisher regarding the use of its published material.

Source data and licence position

Independence and attribution statement

Beacon contains data sourced from Transpower New Zealand Limited, the Electricity Authority Te Mana Hiko, and Stats NZ Tatauranga Aotearoa.

Lumere is independent of these organisations. Lumere is not endorsed by, sponsored by, affiliated with, authorised by, or acting on behalf of Transpower New Zealand Limited, the Electricity Authority Te Mana Hiko, Stats NZ Tatauranga Aotearoa, or any related agency.

Transpower New Zealand Limited has not reviewed, approved, endorsed or validated Beacon, Lumere's methodology, or any Beacon ranking, classification, score or recommendation.

Any Beacon analysis, compilation, transformation, calibration, scoring, ranking, classification, commentary or recommendation is Lumere's own work. It should not be attributed to, represented as approved by, or relied on as advice from any source data provider.

Source data is used on an "as published" basis. Beacon does not reproduce Transpower's tools, does not represent Transpower's operational view, does not represent confirmed connection capacity, and does not replace direct engagement with Transpower, the Electricity Authority, Stats NZ, network companies, generators, consultants or other relevant parties.

Beacon does not make Transpower's original source files available for download as source files. Where Beacon displays, reproduces or adapts selected Transpower connection pipeline information, it does so with attribution and subject to the important notice above.

Where applicable, source data is used under the Creative Commons Attribution 4.0 International Licence. CC BY 4.0 permits sharing and adaptation, including commercial use, provided appropriate credit is given, a link to the licence is provided, changes are indicated, and the use does not imply endorsement by the licensor.

Attribution

Beacon credits its openly licensed sources in the form each licensor specifies.

• 1. Transpower interconnection branch-capacity data. Source: Transpower New Zealand Limited. Published under the Electricity Industry Participation Code 2010 and reusable under the Creative Commons Attribution 4.0 International Licence, with attribution to Transpower New Zealand Limited, as confirmed by Transpower in August 2026. Used as indicative asset-capability information only.
  Licence: Creative Commons Attribution 4.0 International
• 2. Transpower connection pipeline data. Source: Transpower New Zealand Limited. Reusable under the Creative Commons Attribution 4.0 International Licence, with attribution to Transpower New Zealand Limited, as confirmed by Transpower in August 2026. Used as indicative contextual and screening information only.
  Licence: Creative Commons Attribution 4.0 International
• 3. Transpower Envision Opportunities information. Source: Transpower New Zealand Limited. Used with attribution as an indicative guide only. Beacon's derived headroom, N-1, scoring and classification outputs are Lumere-derived screening outputs and are not official Transpower outputs.
• 4. Transpower geospatial network data. Source: Transpower New Zealand Limited open geospatial data. Licensed under CC BY 4.0 where stated by Transpower for the relevant dataset, with attribution to Transpower New Zealand Limited.
  Licence: Creative Commons Attribution 4.0 International
• 5. Electricity market data. Source: Electricity Authority Te Mana Hiko / EMI. www.emi.ea.govt.nz provided by the Electricity Authority (New Zealand). Licensed under CC BY 4.0 unless otherwise indicated by EMI. Includes nodal pricing, reserves, metered generation output, embedded generation and the dispatched generation plant register. www.emi.ea.govt.nz
  Licence: Creative Commons Attribution 4.0 International
• 6. Regional boundaries. Source: Stats NZ Tatauranga Aotearoa. Regional Council 2025. Licensed under CC BY 4.0.
  Licence: Creative Commons Attribution 4.0 International

Licensing positions last reviewed: August 2026, following written confirmation from Transpower New Zealand Limited.

How we weight the queue

Not every megawatt in the connection queue carries the same likelihood of proceeding. A consented project under construction is materially more likely to proceed than an early-stage or speculative project. Before Beacon measures queue pressure, it scales each queued project using its consent status and delivery stage together.

Early-stage, speculative and on-hold projects are discounted more heavily. Consented and delivery-stage projects count at higher weightings. The result, which Beacon calls probability-weighted MW, feeds the platform's saturation and headroom screens. This means a location is not marked as constrained solely on the basis of project interest that may never build. The exact weightings are Lumere's own calibration and are used as a screening device only. They are not a prediction that any project will or will not proceed.

Reading a node

A node is a grid connection point. The screening question at a node is simple. Does the public data suggest there may be headroom worth investigating further?

Beacon compares the node's probability-weighted generation queue against Beacon's headroom screen, a worst-case N-1 winter indicator derived by Lumere from publicly available Transpower Envision Opportunities information. This is a conservative screening indicator, not confirmed connection capacity, operational grid capability, or a substitute for project-specific feasibility studies.

Non-generation load is shown alongside as context and is not folded into the number, because its true effect on headroom cannot be settled without power-flow modelling and more detailed system analysis. A node-level read is also incomplete if the export corridors leaving it are already constrained, so Beacon considers the constraint of the most limiting evacuating corridor. Nodal price is shown as a ranking, not a gate. Three-year average prices across New Zealand sit inside a relatively narrow band, so price helps separate otherwise similar nodes rather than deciding the call on its own.

• Opportunity. Public data suggests headroom worth investigating. The weighted queue sits comfortably below Beacon's headroom screen, and its export corridors are not the binding screening constraint.
• Consider. The location appears to be tightening, has material queue pressure, or has a read softened because Beacon holds only part of the relevant data. Worth a closer look, but the specifics need to be tested.
• Avoid. Public data suggests limited remaining headroom, or a likely corridor constraint. Further development may require deeper technical investigation, changed assumptions, or network investment.
• Local distribution. Beacon does not hold published capacity or pricing for this point. Usually this is a connection or distribution asset, or a new connection not yet represented in the annual capacity information. Shown for orientation, with no call attached.

Reading a corridor

A corridor is a transmission route between nodes. The screening question shifts from connection headroom to export headroom. Does the published branch-rating data suggest this route has export capacity worth investigating further?

Corridor capacity is based on Transpower's published interconnection branch rating, the Winter Continuous MVA, converted to MW at a standard power factor. It is a transparent thermal screen, not a security-constrained transfer study and not confirmed available transfer capacity.

Beacon applies the same probability-weighted queue methodology, then looks for the binding span. Where a corridor runs as parallel circuits, Beacon uses the most-loaded circuit as a conservative screening constraint for the route. On long multi-segment lines, such as Benmore to Islington, the most constrained segment along the chain sets the headline. Beacon flags that bottleneck branch so users can see which span appears to be the screening constraint.

• Opportunity. Public data suggests the route has export headroom worth investigating. The weighted queue is well within the screening capacity of the most-constrained span.
• Consider. The route appears to be tightening. Either overall queue pressure is rising, or a single branch appears to be the screening constraint even though the wider route still has some headroom.
• Avoid. Public data suggests the route is heavily constrained or oversubscribed on Beacon's screening method. Exporting more may require further technical assessment, changed assumptions, or network investment.
• Local distribution or display-only. Connection assets, tee-junctions and HVDC links carry no published interconnection capacity in Beacon's source data, so they appear for orientation only, with no call.

Bands and data quality

Every figure is shown as a green, amber or red traffic light. Price and spread are banded against the national distribution, so the bands do not drift simply because the market rises or falls. Saturation uses Lumere's own screening thresholds.

Each call also carries a data-quality band of High, Medium or Low. It reflects how complete Beacon's inputs are for that location, not how attractive the opportunity is. At a node, data quality rises with a matched pricing point, a usable price history and a suitable capacity indicator. At a corridor, it reflects how cleanly the pipeline maps to branches, branch coverage, and how well the endpoint prices agree.

A Low data-quality band can soften an Opportunity to Consider, but it never manufactures an Avoid. Data quality flags how far to trust a call. It does not invent a negative signal.

What is covered, and what is not

Beacon covers Transpower's published interconnection network, the high-voltage backbone with published capacity information. Nodes and branches with sufficient public inputs carry a Beacon call.

Beacon does not cover connection assets or radial feeds from the backbone to local grid exit points where equivalent public capacity information is not published in the same way. Those lines are drawn for orientation only, with no investment signal. HVDC links and infrastructure deviations are shown the same way. Where users see Local distribution or a display-only line, it reflects missing, unpublished or unsuitable input data, not a representation that no operational issue exists.

The battery arbitrage read

The arbitrage opportunities lens reads each node as a potential battery site using historical market data, with no forecasts. There are six steps. The sixth, added in July 2026, is what part of the day makes the gap, because a gap made by expensive evenings behaves differently as more generation connects than one made by cheap middays.

• The spread. From 36 months of half-hourly EMI prices, Beacon measures the daily price gap at each node, the gap between the cheapest and dearest hours of a typical day, matched to a battery run-time of 2 hours and 4 hours. This approximates the historical gross price spread available to a battery cycling once a day, before project-specific costs, efficiency, degradation, dispatch strategy, availability or constraints.
• The node's own signal. Spreads rise and fall nationally, so Beacon subtracts each island's median to isolate what is specific to this node, that is whether it has historically been a stronger or weaker place to buy low and sell high than its peers.
• The trend. The headline question is whether this node's edge is widening or narrowing. Beacon measures this from the last 12 months, using three-month medians at each end of the period so a single month cannot create a false move. The three-year trend is shown beside it as backdrop. When both point the same way, the move appears more durable. When they disagree, the card says so. A node whose year varied month to month is flagged rather than ranked.
• Siting. The daily price gap is not the whole case, so Beacon surfaces grid-strength context, storage competition, and reserves. Grid strength is shown as context for how difficult or costly a battery may be to connect. Storage competition identifies battery capacity already operating or queued at the node. Reserves are shown as a second island-priced revenue stream that every new connecting battery competes for.
• What makes the gap. Beacon splits the daily gap into the evening climb, which is how far the dearest hours sit above the day's average, and the overnight dip. The evening climb is the part that differs between places, and it is largest where there is little local generation and power has to be brought in at peak. That matters because an evening gap narrows as local generation is built, while a midday gap widens.
• First mover. Where a standalone battery is already operating at a node, Beacon steps the band down by one, for example strong to moderate, because an operating standalone battery may already occupy a strong connection position and an early reserves-market position. A new project may therefore be a second mover competing for what remains. Beacon applies this when the operating standalone battery is 50 MW or more and names it on the card. Batteries co-located to firm a solar farm do not trigger this adjustment, since they are not chasing the standalone daily price gap in the same way. This is a screening signal, not a bankability verdict.

The green, amber or red call is set by the daily price gap ranked across New Zealand nodes, then stepped one band where the node's premium is widening or eroding faster than its region. The colour therefore reflects both how large the historical spread is and whether the node's edge appears durable. Where a spread ranks high but the node's own price record is thin, the band is capped one level down as a guardrail and the card says so. A thin-data node never shows green, and a node with a standalone battery already operating steps down.

A strong BESS node has a wide historical daily spread, a durable or widening node-specific trend, workable grid-strength context, apparent connection headroom, and no first mover already operating. Beacon ranks locations on what three years of market data show. It does not publish forward projections. The signal is historical evidence, not a forecast.

Price erosion lens (solar & wind)

The price erosion lens is a measured, backward-looking screen of what solar and wind output has historically earned at each node, whether that value is eroding, and where export constraint risk may be accumulating. It is not a forecast or a bankability tool.

The price erosion lens addresses an early screening question. Has solar or wind output historically earned enough at this location, and is that value showing signs of erosion? Two different mechanisms can erode a renewable project's value, and Beacon keeps them separate because it can measure one and only screen the other.

• Price erosion. Output may be worth less. Wind and solar often generate at the same time as similar resources across a wide area. When that happens, prices can be lower during the very hours those projects produce. A wind or solar MWh can therefore be worth less than the average MWh. Beacon measures this as the capture factor, the average price the technology historically earned divided by the average price across all hours. A capture factor of 100% means the technology earned the average price. Below 100% indicates a discount, or cannibalisation. The trend shows whether that discount is deepening.
• Export constraint. Output may be blocked by network limits. Where local or export capacity is constrained, generation may be reduced through dispatch, operational constraints, outages or project-specific arrangements. Beacon screens this from the grid intelligence saturation and headroom indicators. It is an indicator of risk, not a measurement of output actually constrained off, and the connection card carries it in full.

Beacon uses measured data wherever it reasonably can. Wind capture is measured from operating wind farms' actual metered output multiplied by the price at the relevant node. Both wind and solar have historically earned less than the average price in the measured record, and both shares have been falling as more of each technology connects. Because only a few solar farms operate today, solar leans more heavily on what operating solar across New Zealand earns today, adjusted for each node's modelled daily price pattern, and shifts toward fully measured treatment as the fleet grows. Beacon also checks the observed relationship across operating farms. Wind capture has historically fallen as more wind connects nearby, supporting the view that value erosion is observable in the historical data, not just assumed. Solar points in the same direction more weakly while its fleet is still small. The current measured levels and trends for each technology are published on the monthly briefing rather than here, because they are re-derived every month and this page describes the method, not the result.

Beacon removes obvious noise on purpose. A new farm's first months of commissioning can generate little and erratically, so Beacon drops them and measures only steady operation. Otherwise, a ramp-up period could look like a sudden value crash that is not real.

Price erosion leads the call. Today's capture level sets the band and a confident measured erosion trend can move it one step worse. Where the export path reads full, the call is escalated one further step and the reason is named on the card. The export path can never set the call on its own, because it is screened rather than measured and the connection card already carries it. Each node is flagged as measured, where an operating farm is present, or analogue, where Beacon estimates from the nearest relevant fleet, so users can see how direct the evidence is.

Each place is measured on its own. Beacon used to compare a node with the average of the other farms in its island, but so few solar farms have run long enough to measure that a node was being compared with a handful that included itself. Island and national figures are still shown, as context beside the local number rather than subtracted from it. The current count of measured farms is on the monthly briefing, which is where figures that move month to month belong.

This is a measured, early-warning screen, not a revenue forecast or a bankability study. Use it to decide where a deeper forward study may be worth doing.

Screening tool, not a bankability or forecasting product. All figures observed and backward-looking; leading indicators are derived from announced or measured facts.

What a battery beside the generator would have added

Beacon takes the prices that really happened at that place, half hour by half hour, over the last year. For each day it moves output out of the cheapest hours the generator was running and sells it back in the dearest hours of the same day. The difference is what a battery would have added for every megawatt hour generated. Where a farm operates, its own record is used. Where none does, a typical national generation pattern is used instead and the card says so.

Four hours is used as the headline because that is where the gains flatten. A day has one cheap block and one dear block to move power between, so once storage can cover that shift, more hours add very little.

The battery is operated as if the day's prices were known in advance, so treat every figure as a best case rather than an expectation. It leaves out wear on the battery, network and connection charges, power used by the site itself, and any money earned from reserves or other services. It also ignores the limit on how much can be exported, which if anything understates the benefit where the grid is tight.

What crowding does to earnings

Beacon compares regions with each other at the same moment. Comparing over time does not work, because a wet year or a fuel shortage moves every region together. Set side by side, the regions with the most of one technology already built earn the least for it.

Wind has been built out for longer than solar, so it shows the clearest picture of what crowding does. Across the wind farms with enough history to measure, each additional 100 MW built in a region has come with a measurably lower share of the average price earned. The regions with the most wind built earn the least for it, and the emptiest regions earn the most. Solar works the same way, in the middle of the day rather than when it is windy. The measured slope, the sample it is drawn from and the regional ranking are published on the monthly price erosion page, because all three move as farms are built and as each new farm passes twelve months of operation.

This is a pattern across places, not proof that one causes the other. Regions differ in ways Beacon does not measure. Wind and solar also push prices down at different times of day, so the wind figure shows how the effect works rather than a number to apply straight to solar. Too few solar farms have run long enough to measure for solar's own line to be drawn yet; the monthly briefing states how many there are and how many more are needed before Beacon will publish it.

What makes the daily price gap

A battery earns on the gap between the cheapest and dearest hours of the day. Beacon splits that gap in two. The evening climb is how far the dearest hours sit above the day's average, and it is the part that differs materially from one place to the next. The other part is the overnight dip, and it is close to the same everywhere, because the shape of the day is set across a whole island rather than at one node. The current spread of evening climb across the regions is on the monthly briefing, where it is restated each month against fresh prices.

Beacon also watches which part of the day is cheapest. Today it is overnight almost everywhere. If enough solar connects, the cheapest hours move to the middle of the day, and the best places for a battery move with them. Twelve of the forty eight half hours in a day sit in the midday window, so picking hours at random would put a quarter of them there. Days with almost no price movement are left out, because on those days the cheapest hours are close to random.

The middle of the day can also go cheap because the lakes are full or demand is low, so a rise in this number is a reason to look rather than proof that solar has changed the market.

How we measure change (What's Changed)

Each monthly briefing ranks the nodes and corridors that genuinely moved. A move is measured over the last 12 months and never from a single month at each end. Beacon takes the median of the first three months and the median of the last three months of the window and compares those, so one extreme month, price spike or dry snap cannot create a false year-long move.

A node only enters the ranked lists if its year actually trends one way. At least 60% of its month-to-month steps must agree in direction, with at least ten months of usable data behind them. A node that swings back and forth is flagged as varied month to month and sits out of the rankings until its story settles.

Small moves are treated as noise, not news. A move must clear $3/MWh to rank, or $2/MWh for the watch list.

Where a material move happens at a node whose price record is statistically thin, Beacon neither presents it as a clean signal nor hides it. The briefing counts those suppressed moves and discloses the count, so the picture is honest in both directions.

Regional editions rank with the same method inside the region and show the national leaders alongside, so a quiet region is not mistaken for a quiet country.

Hydro years and the price cycle

New Zealand's wholesale price level is set substantially by hydro storage. A dry year can lift prices across an island for months at a time. A wet year can suppress them. The current three-year pricing window includes a clear dry period, and its imprint is most visible in South Island price levels. Any tool built on recent market data carries that cycle, so Beacon is explicit about where it shows up and how it is handled.

Absolute dollar figures, such as a node's three-year average price, are shown as they were earned. Beacon does not adjust them for hydrology because it holds no storage or inflow model, and a hydro-corrected price would be a modelled forecast, which Beacon deliberately does not publish. Instead, price bands are set against the national distribution, so a wet or dry year moves the whole distribution together rather than automatically flipping individual nodes between bands.

On the price erosion card there is one deliberate exception. The dollars a solar or wind project would earn are shown on the node's last twelve months, not its three-year average. This is because the capture rate they are multiplied by is measured over the operating fleet's recent record, and a recent rate paired with a three-year price level that still carries the 2024 dry year produces an earned figure no farm has ever seen. Matching the price window to the capture window keeps the estimate honest. The three-year average is still what the investment rating screens on, because there the number is relative and a common price level cancels out across every node at once.

The node-specific signals are measured relative to the island. Price basis is the node against its island average. The battery spread edge is the node's spread against its island median. Solar and wind capture is read against the island reference. A hydro swing moves the whole island broadly together, so the island-relative read removes much of the common distortion. That is the deliberate design. Comparing a node to its island is Beacon's hydrology normalisation.

Trends are protected from single-month distortion. Twelve-month moves are measured between three-month medians at each end of the window, so one extreme hydro month cannot create a false trend, and a node whose year varied month to month is flagged rather than ranked.

Where a dry year widens price dispersion enough to weaken statistical confidence, the affected reads carry a lower confidence band, and material moves at those nodes are disclosed as suppressed rather than presented as clean signals.

The honest bound is that island-relative measures remove much, but not all, of the hydro effect. Dry years can also change the shape of the day and shift the balance between the islands across the HVDC link. That residual is part of why Beacon is a screening tool, and why every read should be tested against the user's own view of market conditions before capital is committed.

The investment rating

The investment rating is a relative grade, from 0 to 100, of how good a connection point is to develop, scored against every other connection point in New Zealand. It is built from four measured signals. These are how much firm capacity there is to export from the node, how congested the road out is, how well generation captures the price in the hours it actually sells, and the nodal price level. The four are weighted by Lumere’s calibration, with the grid signals and price-capture carrying the bulk of the weight and the price level a lighter touch; the exact weighting is part of Lumere’s methodology.

The rating leads on the best technology you could actually build at the node, wind or solar, whichever is stronger, because a developer builds their best option rather than the average of their options. A node that can do both well earns a small premium for that flexibility; it is never penalised for having a weaker second choice.

Crucially, the rating is not a single number. It re-rates for the size of the project you are screening. A connection point that comfortably hosts a 20 megawatt project can be constrained at 100 megawatts if it cannot firmly export that much, so the same node can read as an opportunity at one size and be set aside at another. Every rating is gated by the firm export capacity, the megawatts that can genuinely leave the node, not by the raw size of the connection.

A separate 2028 view carries today’s development pipeline forward. Committed projects are firmed to their full size and consume headroom, and published network works add headroom back. The most important forward effect is price-capture erosion. When a whole island fills with wind, the fleet peaks together and drives the price down in exactly the hours a wind farm sells, so its capture erodes island-wide as more is built, and steady baseload demand such as a smelter or a data centre cannot rescue those windy-hour prices. Solar behaves the same way at midday. The rating reads this coincidence directly rather than assuming local demand soaks it up.

Competition and price-erosion pressure are read across six electrical zones, upper, central and lower in each island, rather than administrative regions, because generation competes for price and for the way out electrically, not by geographic boundary. Within a zone the dominant sub-area is named, so a concentrated build-up is not diluted and a spillover one is not understated.

The achieved price a node is shown to earn is measured capture applied to a hydrology-normalised forward hub price, with a locational adjustment for how the node prices against its island hub. The normalisation is deliberate: it removes the distortion a single wet or dry year puts into a recent average, so the figure reflects a through-cycle level rather than last year’s hydrology. It is a screening estimate, not a forecast of what a project will earn.

Batteries are shown as a separate trajectory read, not as part of the build rating. Merchant arbitrage in the New Zealand market does not yet stack up as a standalone return, so a battery never sets a node’s headline grade, though the same island glut that erodes generation capture is what widens the spread a battery would live on.

If a component cannot be measured the weights renormalise and the confidence drops a notch; with fewer than three components Beacon publishes no grade and says so. Grades re-rank every month as the pipeline moves, and every move is reported in What’s changed.

What Beacon is not

How to read the rating, and where it stops

The rating is a grade, not a return. It ranks a connection point against its peers on grid conditions; it is not an internal rate of return, and a high grade is not a promise of a bankable project. The absolute figure to take away is the firm export capacity shown on every node, because that is the number a project actually lives within.

It measures the grid, not the resource. The score reads headroom, congestion, capture and price. It does not model the wind or solar resource itself, so a southern node can show a workable solar grade on favourable grid conditions even where the irradiance is genuinely weaker than the north. Weigh resource quality separately.

It does not price the cost of connecting. A well-rated node can still carry a large connection or network-reinforcement cost. The rating screens where the grid is favourable to develop, not what it costs to plug in there.

It assumes firm export at full output. The size gate treats your full project as needing to export firmly at all times. A developer willing to accept some spill may find a node workable above its firm limit, so read the size gate as the firm case rather than an absolute ceiling.

It caps project size by connection voltage. A busbar can only physically host a project so large before the voltage itself rules it out, so the rating applies a ceiling by voltage, taken from what actually connects in the New Zealand queue rather than a flat rule: roughly 150 megawatts at 33 kV, 180 at 66 kV and 250 at 110 kV, with 220 kV and above effectively uncapped for the sizes screened. A size beyond its voltage is ruled out of the rating entirely, carrying no score, rather than graded low. Below that ceiling the firm export check does the real work, so the same size can fit at a strong bus and be set aside at a weak one. These are defensible screening defaults, not a substitute for a connection study.

The forward view rests on a probability-weighted pipeline. Queued projects are weighted by how far they have progressed, so a consented project counts for more than an early-stage one. As projects advance or new ones enter, the forward ratings tighten automatically the following month; nothing here is pinned to a single snapshot. The rating is not knife-edge on those weights. Flexing the whole pipeline up or down by 20% moves a node’s rating by roughly a single point and does not reorder the leaders, because a uniform shift moves the whole board together. The movement that matters comes from individual projects entering, progressing or dropping out, which is exactly what the monthly re-rate captures and reports.

Firm export is checked against the wires out wherever a published corridor figure exists. Where a low-voltage bus has none of its own, it inherits the lower of its own capacity and its substation’s route-verified export, since every bus at a site shares one road out, so an unchecked bus figure can no longer set the number on its own. A connection point with no route-checked bus anywhere is shown as an estimate and held off the ranked boards, enforced by an automated check. Corridor coverage is expanding month on month.