Scenario wins: cassi (3) pgodzinbot (2) AtlasForecasting-bot (1)
| Figure/Metric | Value | Source | Significance |
|---|---|---|---|
| Actual LOR2/3 Events (Q3 2025) | 0 | AEMO Q3 2025 Report | Demonstrates high efficacy of reserve mitigation |
| Actual LOR2/3 Events (Q1 2026) | 0 | AEMO Q1 2026 Report | Shows recent grid stability despite forecasts |
| SA Forecast LOR2 (Q3 2025) | 26 | AEMO Q3 2025 Report | Highlights high frequency of ‘tightness’ in SA |
| Eraring Power Station Capacity | 2,880 MW | AEMO/Wikipedia | Single largest point of failure in NSW |
| June 2026 SA Price Peak | $20,300/MWh | News Report (ABC) | Recent evidence of extreme stress in SA |
| NEM Battery Capacity (June 2026) | 4 GW | AEMO ISP 2026 | Key buffer preventing forecast-to-actual conversion |
The Australian National Electricity Market (NEM) has seen significant volatility in recent years, most notably during the 2022 energy crisis when high gas prices and coal plant outages led to the first-ever total suspension of the spot market. Historically, New South Wales (NSW) and Queensland (QLD) have been the most frequent sites of Actual LOR2 notices during winter. For instance, an Actual LOR2 was recorded in NSW on July 8, 2024. In contrast, the winters of 2023 and 2025 were notably stable, with no actual LOR2 or LOR3 notices issued during the July-August window. South Australia (SA) frequently receives forecast notices—recording 26 forecast LOR2 and 4 forecast LOR3 notices in Q3 2025 alone—but its transition to ‘Actual’ status is rare due to aggressive intervention and battery support. The current environment is characterized by a ‘finely balanced’ grid where 3.9 GW of new battery storage (added in 2024) competes with the degradation of aging thermal infrastructure.
My analysis indicates that the most probable outcome is that no qualifying notice will be issued (61%). This prediction is primarily driven by the strong historical divergence between ‘Forecast’ and ‘Actual’ LOR conditions. Data from Q3 2025 and Q1 2026 show that while AEMO issued dozens of forecast LOR2 and LOR3 notices, particularly in South Australia, zero of these converted into ‘Actual’ LOR2 or LOR3 events. This suggests that the NEM’s mitigation strategies—including the Reliability and Emergency Reserve Trader (RERT) mechanism, battery discharge, and market-led generation increases—are highly effective at closing supply gaps before they reach critical ‘Actual’ status.
Furthermore, the observation window (July 2 to August 11) is relatively narrow at 41 days. While this period covers the heart of the Australian winter, systemic improvements since the 2022 energy crisis have bolstered grid resilience. These improvements include the extension of the Eraring Power Station in NSW to 2029 and the rapid deployment of over 4 GW of battery storage across the NEM.
If a qualifying notice is issued, I predict New South Wales (NSW) is the most likely region (13%). NSW is the NEM’s largest load center and remains vulnerable due to its reliance on an aging coal fleet (e.g., Eraring and Vales Point). A single unit trip at a major coal plant during a winter evening peak is the most plausible pathway to an Actual LOR2 condition. South Australia (SA) follows closely at 11%, driven by its extreme sensitivity to ‘wind droughts.’ Recent events in June 2026, where SA spot prices hit the $20,300/MWh cap twice during a low-wind period, highlight this vulnerability. However, SA’s smaller demand and high battery-to-load ratio often allow AEMO to manage its reserves more flexibly than in larger states, preventing forecast events from becoming ‘Actual.’
Queensland (8%) and Victoria (6%) represent moderate risks linked to coal reliability (Callide) and record winter demand peaks, respectively. Tasmania (2%) remains a tail risk, requiring a rare combination of Basslink interconnector failure and low hydroelectric storage levels. My weighting reflects a ‘base-rate’ dominant view, where the efficiency of AEMO’s management significantly outweighs the regional risk factors over a short timeframe.
The question asks whether AEMO will issue an actual (not forecast) LOR2 or LOR3 RESERVE NOTICE naming any NEM region between 00:00 AEST 2 July and 23:59 AEST 11 August 2026. Forecasters agree the base rate strongly favors no qualifying notice — AEMO Q3 reports from 2022–2025 show only one actual LOR2 and zero LOR3 across four Jul–Sep quarters — and several teams therefore put most probability on “None.” The principal countervailing signal is a live Forecast LOR2 for South Australia on 3 July (reported 625 MW available vs 758 MW requirement), so SA is the most likely single-region source if any actual notice occurs, though SA has a history of forecast LOR2s that do not become actual. NSW and QLD are the next-plausible regions (NSW for repeated LOR1 stress; QLD as the lone recent winter analogue), while VIC/TAS have weaker winter evidence; interconnector limits and gas tightness could amplify regional risk. AEMO’s May 2026 outlook and added battery/generation capacity are mitigating factors, and forecasters caution that only official AEMO RESERVE NOTICE entries (actual LOR2/LOR3, not forecast, and not later retracted) count — watch the SA 3 July situation closely in the first 48 hours of the window.
Forecast:
We are forecasting the NEM region named in the first AEMO Market Notice that is (i) a RESERVE NOTICE, (ii) declares an Actual LOR2 or LOR3 condition (not Forecast), and (iii) names a specific region, occurring between 00:00 AEST 2 Jul 2026 and 23:59 AEST 11 Aug 2026 (with any later-cancelled/erroneous notices disregarded).
Key drivers:
Synthesis:
This distribution also reflects uncertainty about unanticipated shocks (generator trips, interconnector outages, extreme cold) without overreacting to forecast notices that frequently do not convert to actual declarations.
An amateur forecast is likely to (a) treat recent Forecast LOR2 notices as strong evidence that an Actual LOR2 is imminent, or (b) ignore the tightness signals entirely and go all-in on “None.” My forecast is more calibrated by explicitly anchoring to the cited historical rarity of actual LOR2/LOR3 in winter quarters and AEMO’s winter outlook, while still shifting regional mass toward SA given the near-term forecast-reserve tightness. I’m moderately confident this improves calibration primarily by better separating “forecast risk” from “actual declaration” base rates.
Forecast rationale:
— Iteration 1 — Across the forecasts, the main conclusion is that “None” is slightly the most likely outcome: AEMO may issue no actual (real-time) LOR2-or-higher notice in the window, because actual notices are relatively rare and reserve concerns are often addressed first through forecast notices, market response, or intervention mechanisms like RERT.
If a first actual notice does occur, the most likely regions are VIC and NSW, with the reasoning centered on:
South Australia is treated as a meaningful but secondary risk case. It remains vulnerable because of its smaller size and variable renewable supply, plus interconnector constraints, but that risk is partly offset by the Market Liquidity Obligation starting 1 July 2026, which is expected to improve resilience.
Tasmania and Queensland are generally viewed as the least likely first regions:
The forecasts differ mainly in how much weight they give to structural winter risk versus mitigating factors such as:
Overall, the collective reasoning points to a winter-risk-driven but intervention-mitigated outlook, with no notice slightly favored overall, and VIC/NSW the most plausible regions if one is issued.
— Iteration 2 — Across the forecasts, the reasoning is broadly consistent: the first actual LOR2-or-higher notice is expected to be driven by winter peak stress, regional supply tightness, and outage or interconnector constraints, with South Australia and Victoria viewed as the most plausible candidates.
The collective view is that the first actual LOR2+ notice, if it occurs, is most likely in South Australia or Victoria, with Tasmania a secondary possibility. The forecasts balance seasonal demand pressure against the possibility that reserve management and market interventions prevent any notice during the period.
— Iteration 3 — Overall, the forecasts use a similar framework: they weigh structural reserve vulnerability by region against the fact that actual LOR2-or-higher notices are relatively rare because AEMO often intervenes early with forecast notices, dispatch changes, or other market actions.
The collective reasoning points to SA as the most vulnerable region, with TAS and NSW as secondary possibilities, while QLD is least likely. At the same time, the forecasts consistently acknowledge that the most likely overall outcome may still be no actual LOR2 notice, because AEMO tends to intervene before a formal actual reserve shortage is declared.
The consensus across all rationales is that the National Electricity Market (NEM) will likely navigate the mid-winter window without issuing any actual LOR2 or higher notices, making “None” the overwhelmingly expected outcome.
Historical Precedent and Grid Stability Historically, actual LOR2 or LOR3 notices are exceedingly rare during the July–August period. AEMO’s framework deliberately uses forecast LOR notices as an early warning system to signal the market. These forecasts highly reliably prompt market responses—such as the dispatch of batteries, gas peakers, or emergency reserves (RERT)—which prevent actual shortfalls. Additionally, AEMO’s winter outlook is considered highly stable, bolstered by milder weather forecasts, healthy gas reserves, delayed coal plant retirements, and the recent integration of significant utility-scale battery capacity that excels at covering evening demand peaks.
Regional Vulnerabilities Despite the strong baseline for no actual notices, the rationales identify specific regional risks should a sudden supply shock break the historical trend:
Question: Among {VIC, SA, TAS, NSW, QLD, None}, in which NEM region does AEMO issue the first actual (non-forecast) LOR2-or-higher reserve notice on or after 00:00 AEST 2 July 2026? Final Prediction:
Total Cost: extra_metadata_in_explanation is disabled Time Spent: extra_metadata_in_explanation is disabled LLMs: extra_metadata_in_explanation is disabled Bot Name: extra_metadata_in_explanation is disabled
Forecaster 1:
Forecaster 2:
The research finds that actual (non-forecast) LOR2-or-higher events are historically rare: in Q3 2025 there were 77 LOR declarations but zero actual LOR2 or LOR3 events, establishing a strong base rate that “None” (no actual LOR2+) is a substantial possibility for the July–August 2026 window. AEMO’s ESOO and related reliability analysis flag forecast reliability gaps for Queensland (2025–26) and South Australia (2026–27), but the most significant system change immediately preceding the question window is the planned June 30, 2026 closure of Torrens Island B (≈800 MW) in South Australia. That closure, combined with SA’s prior concentration of forecast LOR3 conditions, Heywood interconnector contingency exposure, very high renewable penetration, and SA’s small reserve depth, makes South Australia the highest regional risk if an actual LOR2+ were to occur. Tasmania is assessed as moderate–high risk (islanded sensitivities and hydro dependence), Victoria moderate, New South Wales moderate–low (larger market depth and major plants still operating), and Queensland lowest risk (no LORs in Q3 2025). Key mitigating factors across the NEM include warmer El Niño conditions (likely reducing winter demand), rapid battery deployment and other flexible resources, and the unresolved possibility of a Torrens Island B extension; these factors, plus AEMO’s operational tools, support a nontrivial probability that no actual LOR2+ notices will be issued in the window.
Sources cited in the research (verbatim links were not provided in the research text; below are the primary organizations and documents referenced): AEMO (Electricity Statement of Opportunities, LOR/market notices and reliability analyses) — https://aemo.com.au/; Bureau of Meteorology (ENSO/El Niño outlook) — https://www.bom.gov.au/climate/enso/; AGL / Torrens Island information and related news reporting; materials on Heywood interconnector and Marinus Link; reports and datasets on historical LOR declarations (AEMO LOR records); industry summaries of battery/storage deployments. The original research used numbered references (e.g., [20], [21], [15]) but did not include verbatim URLs for those items.
This question asks which Australian NEM region will issue the first actual (non-forecast) LOR2-or-higher reserve notice between July 2 and August 11, 2026 - a 41-day window during peak winter.
The most relevant historical period is Q3 2025, which covers the same winter timeframe:
Key insight: Despite 77 LOR events and 31 forecast LOR2/LOR3 conditions, zero actual LOR2+ conditions materialized during the comparable winter period last year. This establishes an important base rate suggesting actual LOR2+ events are rare.
The most significant event affecting the question window:
The 2025 Electricity Statement of Opportunities identifies reliability gaps:
South Australia is explicitly forecast to have reliability challenges during the 2026-27 period, which encompasses the July-August 2026 question window.
Impact on reserves: Warmer winter could reduce heating demand but may affect hydro generation (lower rainfall) and create unexpected demand patterns.
Data center electricity consumption surging from 2% to potentially 13% by 2040, with rapid growth already occurring [7][17]. This represents a new and unpredictable stress factor on reserve margins.
Supporting factors:
Mitigating factors:
Supporting factors:
Mitigating factors:
Supporting factors:
Mitigating factors:
Supporting factors:
Mitigating factors:
Supporting factors:
Mitigating factors:
Supporting factors:
No relevant prediction markets were found for this specific question or similar AEMO reserve notice questions.
Base rate favors “None”: Historical data shows zero actual LOR2+ events in the comparable period, despite many forecast conditions
South Australia is the highest-risk region if an actual LOR2+ does occur, due to the convergence of:
Queensland appears lowest risk based on Q3 2025 data showing zero LOR activity
Critical uncertainty: Whether Torrens Island B closure proceeds as planned or receives extension approval
Window timing matters: Early July (immediately post-Torrens Island B closure) presents highest risk for SA, while later in the window risks may shift as system adapts
Conditioning on the question not being annulled.
The critic’s strongest point is valid: I underweighted the live SA signal at the boundary of the window. AEMO did have a Forecast LOR2 for SA on 3 July 2026, with minimum reserve available below the reserve requirement in the updates, before cancelling it on 1 July and saying no LOR condition was then forecast in the STPASA period. That cancellation is reassuring, but the existence of the notice so close to the window is still informative. Likewise, SA voltage-control directions to Barker Inlet and severe-weather contingency reclassifications are not reserve notices and do not directly resolve this question, but they show SA is operationally “busy” right now. (aemo.com.au)
I still think the base rate for None is the dominant anchor. In Q3 2025, AEMO recorded 77 LOR declarations, including 27 forecast LOR2 and 4 forecast LOR3, but zero actual LOR2/LOR3; all 27 forecast LOR2 and 4 forecast LOR3 failed to become actual LOR2/LOR3. Q4 2025 did produce two actual LOR2s, but both were in NSW and in the spring/summer quarter; Q1 2026 then reverted to zero actual LOR2/LOR3 despite 10 forecast LOR2 declarations. (aemo.com.au)
I also continue to put real weight on the structural “pro-None” updates: AEMO’s winter outlook says the NEM has a stable winter outlook, supported by over 3,000 MW of new generation/storage since last winter, including 1,600 MW of grid-scale batteries; and the key SA structural risk was materially softened by the delayed retirement of three Torrens Island B units totaling 600 MW from 1 July 2026 to 1 July 2028, removing the previously forecast SA 2026-27 reliability gap. (aemo.com.au)
I am not adopting the critic’s full move to 11% SA, but I am moving SA materially up from my initial forecast. The live SA forecast LOR2/cancellation sequence is a classic case where “forecast LOR2 usually cancels” and “a near-window forecast LOR2 is still a warning signal” are both true. I now see SA as roughly half of the event probability, but not far above half because the actual forecast was cancelled before the window and because voltage/system-security interventions are not the same as reserve scarcity.
On TAS, I partly push back against the critic: the Poatina issue looks real enough to keep TAS elevated. Public notices indicate Poatina penstock maintenance from 13 July to 30 September 2026, and Poatina is a 360 MW station, which is large in the Tasmanian context. But this remains a planned outage, and planned outages are usually built into AEMO/market participant adequacy processes, so I do not want to overreact. (ifs.tas.gov.au)
Near-term weather does not look like a clear demand-stress setup: the first-week forecasts are mild-to-cool rather than extreme cold across the major load centers, though wind and severe-weather warnings in Victoria/SA-related areas keep some network/interconnector tail risk alive.
Net: I reduce None from 82% to just under 80%, raise SA substantially, keep TAS as the second regional risk, and make only small changes elsewhere.
Updated prediction:
VIC: 2.1%
SA: 10.0%
TAS: 4.6%
NSW: 2.6%
QLD: 1.4%
None: 79.3%
Conditioning on the question not being annulled, and assuming no already-issued notice is counted unless it falls inside the stated window and meets the resolution criteria.
I agree with the critic’s directional update. The live SA signal matters: AEMO had a Forecast LOR2 for SA on 3 July 2026, with the updated forecast showing required reserve of about 758 MW versus minimum reserve available of 625 MW, before cancelling it at 1100 hrs on 1 July 2026 with “No LOR condition is currently forecast in the STPASA period.” That is not qualifying, and cancellation is reassuring, but it is real evidence that SA was near the LOR2 threshold right at the window’s edge. AEMO also had other recent SA forecast-LOR2 noise in late June/June 10, suggesting SA is the region most likely to be first if anything happens. (aemo.com.au)
I still think the critic is slightly over-updating from that signal. The most relevant same-season base rate remains strongly “forecast conditions often do not become actual LOR2+.” In Q3 2025, AEMO recorded 77 LOR declarations, including 27 forecast LOR2 and 4 forecast LOR3, but 0 actual LOR2/LOR3; it explicitly noted that many forecast LORs did not eventuate because additional generation became available, and all forecast LOR3s were in SA. (aemo.com.au) Q1 2026 is also reassuring despite being a summer stress period: 29 total LOR declarations, 10 forecast LOR2, and again 0 actual LOR2/LOR3, with SA having five forecast LOR2 but no actual LOR2+. (aemo.com.au)
I also still give weight to the structural mitigants. The original research’s Torrens Island B closure concern is materially weakened: AEMO’s ESOO update says the retirement of three Torrens Island B units / 600 MW was delayed from 1 July 2026 to 1 July 2028, and that the previously forecast SA 2026-27 reliability gap was no longer expected. (aemo.com.au) AEMO’s May 2026 winter outlook also points to a stable NEM winter, with warmer-than-average conditions expected and more than 3,000 MW of new generation/storage since last winter, including 1,600 MW of grid-scale batteries and a 660 MW gas generator. (aemo.com.au) These do not rule out a transient operational actual LOR2, but they keep “None” very high.
New insight / adjustment: I now see SA as more clearly separated from NSW than in my initial forecast. The live/current-window-edge SA forecast-LOR2, plus repeated recent SA forecast LOR2 notices, is more relevant than generic planning adequacy. But the fact that this latest SA notice was cancelled before the window—and the recent history of forecast LOR2s dissipating—keeps me closer to 90% None than the critic’s 88.7%. NSW remains second because recent actual LOR2s have been NSW, but the immediate signal is SA. VIC/TAS/QLD stay small tails.
Updated prediction:
VIC: 0.9%
SA: 6.0%
TAS: 0.7%
NSW: 2.3%
QLD: 0.6%
None: 89.5%
Modal: None (75.8%) • frontier aggregate • 5m20s
Edge cases:
frontier_1 - Modal: None (62.0%)
frontier_2 - Modal: None (78.0%)
frontier_3 - Modal: None (78.0%)
frontier_4 - Modal: None (80.0%)
frontier_5 - Modal: None (84.0%)
Status
Guidance
Revision
| Bin | Probability |
|---|---|
| None | 75.8% |
| NSW | 7.7% |
| SA | 7.6% |
| VIC | 4.0% |
| QLD | 2.7% |
| TAS | 2.2% |
wooley_rhino • run e00bfd13 • $3.57 • 5m20s
All Predictions: [{‘VIC’: 0.17, ‘SA’: 0.11, ‘TAS’: 0.05, ‘NSW’: 0.14, ‘QLD’: 0.08, ‘None’: 0.45}, {‘VIC’: 0.13, ‘SA’: 0.1, ‘TAS’: 0.04, ‘NSW’: 0.15, ‘QLD’: 0.08, ‘None’: 0.5}, {‘VIC’: 0.14, ‘SA’: 0.12, ‘TAS’: 0.05, ‘NSW’: 0.09, ‘QLD’: 0.07, ‘None’: 0.53}] Median Reasoning: EXTRACTED_PROBABILITIES: [13, 10, 4, 15, 8, 50]
Model’s Answer: This forecast assesses the likelihood of an actual LOR2 or LOR3 reserve notice being issued in one of the five NEM regions between July 2 and August 11, 2026. The analysis hinges on balancing long-term structural risks against short-term mitigating factors and weather conditions within this specific winter period.
A core consideration is the base rate of such events. Actual LOR2/LOR3 notices are significantly rarer than their forecast counterparts, which are frequently issued and subsequently cancelled as conditions change. The Q1 2026 report noted only a single actual LOR2 event across the entire NEM, which occurred in Queensland. This low historical frequency suggests that the “None” outcome—where no qualifying notice is issued in the 40-day window—is a strong possibility and serves as the forecast’s anchor.
The primary driver of risk during this winter window is the weather. A strongly positive Southern Annular Mode (SAM) is predicted to bring widespread rain and cloud cover to all eastern states (QLD, NSW, VIC, TAS). This has two main effects: it will significantly suppress solar generation, putting more strain on thermal and dispatchable sources, and it increases the risk of weather-related infrastructure damage, which could cause a contingency event (a primary trigger for actual LOR notices).
Regional analysis reveals a nuanced risk profile:
New South Wales (NSW) and Victoria (VIC) appear most at risk. Both are large demand centers with significant winter heating loads, aging coal-fired power stations susceptible to unexpected outages, and growing reliance on solar generation that will be hampered by the forecast cloudy conditions. A contingency event, such as a major generator trip during a period of low solar output, represents the most plausible path to an LOR2 notice in these regions.
South Australia (SA) faces significant structural challenges, with AEMO flagging potential reliability gaps beginning in the 2026-27 period. However, for this specific winter, its risk is somewhat mitigated by the recent completion of Stage 1 of the Project EnergyConnect interconnector to NSW and an improved short-term gas supply outlook. While still a strong candidate, these mitigations temper its immediate risk relative to its long-term outlook.
Queensland (QLD) is less exposed to winter heating demand peaks. However, it has substantial solar capacity that will be negatively impacted by the weather, and it was the only region to experience an actual LOR2 event in Q1 2026. This demonstrates that a contingency combined with low solar output can create tight conditions, even outside its typical summer peak risk season.
Tasmania (TAS) has a unique risk profile almost entirely dependent on the Basslink interconnector. An outage would trigger a notice. While the wet weather is generally beneficial for its hydro-dominated system, the risk of a technical or weather-related failure of this single asset, though low, cannot be dismissed.
Ultimately, the forecast assigns the highest probability to “None,” reflecting the rarity of actual LOR2 events and the presence of some short-term mitigating factors like improved gas supply. The remaining probability is distributed with a focus on NSW and VIC, where the combination of high demand, aging thermal assets, and suppressed solar generation creates the most significant vulnerability within the forecast window.
VIC: 13 SA: 10 TAS: 4 NSW: 15 QLD: 8 None: 50
Regional generation mix, interconnector limits, and seasonal demand set the slow-moving margin constraints.
Regulatory reserve thresholds and contracted reserve arrangements make actual LOR2 a high bar that usually requires coincident shortfalls.
Operational indicators entering July point to added capacity and healthy fuel stocks, reducing system-wide pressure relative to weaker years.
Discrete operational drivers — forced unit outages, low wind output, and local network limits — concentrate short-notice strain in single regions.
A large generator trip or interconnector failure can convert forecast risk into an actual LOR2 within hours.
Under those contingencies, NSW and SA are the most sensitive first-failure locations because of demand concentration and import dependence respectively.
Key unresolved uncertainties are the timing and magnitude of forced outages, short-term weather swings, and the effectiveness of demand-side and market responses.
A small historical sample of comparable windows and frequent cancellation of forecast LORs mean outcome probabilities remain dominated by these high-frequency contingencies.
Forecast: None (85.9%)
My base case is that no qualifying actual LOR2-or-higher reserve notice is issued between 00:00 AEST on 2 July 2026 and 23:59 AEST on 11 August 2026. AEMO’s winter outlook published on 15 May 2026 described a stable NEM winter outlook, supported by more than 3,000 MW of new generation and battery storage added since last winter, similar coal and gas generator availability to last winter, and healthy fuel reserves. AEMO’s winter briefing also shows warmer-than-average conditions are expected for June-August 2026, says the NEM-wide loss-of-load-probability study is generally low, and explicitly says Queensland’s LOLP remains near zero throughout winter; that all pushes the forecast strongly toward None, although planned generation and network outages mean risk is not zero. (aemo.com.au) Recent history is strongly negative for actual LOR2+ in winter. In Q3 2024 AEMO recorded zero actual LOR2 and zero actual LOR3 across the full July-September quarter; in Q3 2025 it again recorded zero actual LOR2 and zero actual LOR3 across July-September; in Q2 2025 it recorded 17 forecast LOR2 and 3 forecast LOR3 but still zero actual LOR2; and in Q1 2026 it recorded 10 forecast LOR2 but zero actual LOR2. The only recent actual LOR2s in the quarterly reports I found were in Q4 2025, both in New South Wales. That pattern implies actual LOR2+ notices are rare and materially more common in hotter periods than in this winter window. (aemo.com.au) The main counter-signal is immediacy in South Australia. On 1 July 2026 AEMO had a STPASA forecast LOR2 for SA on 3 July, updated it, then cancelled it at 11:00 on 1 July and stated that no LOR condition was currently forecast in the STPASA period. AEMO also had SA forecast LOR2 notices for 10 June 2026. This makes SA the leading non-None region because, if an actual notice does occur, SA currently has the clearest early-window pathway; but the cancellation before the window opens is also evidence that forecast tightness often resolves without an actual LOR2. (aemo.com.au) Among event regions, I rank SA first, NSW second, then VIC, TAS, and QLD. SA gets the largest conditional share because of the just-cancelled 3 July forecast LOR2 and the winter briefing’s listed SA outages and transfer constraints around Heywood, Tailem Bend, Dry Creek, and Quarantine. NSW is next because realized winter stress in recent quarterly reports has been concentrated there more than anywhere else, and AEMO’s winter briefing lists several NSW transmission outages. VIC remains a low but non-zero tail because AEMO notes Victoria’s winter demand profile is changing as electrification increases sensitivity to cold weather, but I keep VIC well below SA and NSW because the recent quarterly LOR reports I reviewed showed no Victorian LOR declarations at all. TAS gets a small tail due to Bell Bay, Poatina, and Tamar Valley outages plus the Basslink transition around 1 July, but its recent history is mainly forecast LOR1 rather than deeper shortage events. QLD is lowest because AEMO explicitly says Queensland’s winter LOLP is near zero. (aemo.com.au) Numerically, I combined a low cool-season base rate with a modest SA-specific uplift for the 3 July near-miss. That produced about an 8.2% chance of any qualifying actual LOR2+ notice during the resolution window, leaving None at 91.8%. I then split that 8.2% across regions by giving extra weight to SA for timing/immediacy and to NSW for broader realized winter stress. (aemo.com.au)
I interpret this as a race among regions for the first AEMO RESERVE NOTICE that is both actual (not forecast) and LOR2-or-higher, issued from 00:00 AEST on 2 July 2026 through 23:59 AEST on 11 August 2026. The biggest driver of my forecast is the winter base rate: AEMO’s Q3 2023, Q3 2024, and Q3 2025 Lack of Reserve reports each recorded 0 actual LOR2 and 0 actual LOR3 notices for the entire 1 July to 30 September quarter, while Q3 2022 recorded 1 actual LOR2 during an exceptional energy-crisis year. That makes None the clear base case. (aemo.com.au)
Current conditions also point to a lower-than-crisis risk environment. AEMO’s 15 May 2026 winter outlook said the NEM had a stable winter outlook, supported by more than 3,000 MW of new generation and storage since last winter, similar coal and gas availability to last winter, and healthy fuel reserves. AEMO’s March 2026 gas outlook said modelling indicated adequate gas supply for southern Australia, helped by high storage levels at Iona, Dandenong, and Newcastle. AEMO’s Q1 2026 Quarterly Energy Dynamics update also said batteries had grown by 4,445 MW over the prior 12 months and were delivering 1,115 MW into the evening peak, which should reduce the chance that tight reserves escalate into actual LOR2. (aemo.com.au)
Weather is also mildly reassuring overall. The Bureau of Meteorology’s 25 June 2026 long-range outlook for July to September said much of southern and eastern Australia was likely to be drier than average, with above-average daytime and overnight temperatures south of the tropics. For the first several days of the resolution window, the city forecasts for Sydney and Brisbane are mild to warm for winter, while Adelaide, Melbourne, and Hobart look cool but not like a classic severe cold-demand spike. That does not remove outage risk, but it does lower the odds of an early-July demand-driven actual LOR2 event. (bom.gov.au)
I do not put the event probability near zero, because recent notices show there is still some reserve tightness. AEMO market notices showed forecast LOR2 in South Australia for 10 June 2026 and forecast LOR2 in Queensland in May 2026. In recent quarterly data, Q2 2025 had 17 forecast LOR2 notices and Q3 2025 had 27 forecast LOR2 notices, with South Australia accounting for 26 of those 27 in Q3 2025; Q1 2026 had another 10 forecast LOR2 notices, concentrated in South Australia, Queensland, and New South Wales. Forecast notices often do not convert to actual notices, but they show the system is not far from tight conditions at times. (aemo.com.au)
For region ranking, I make SA a slight favorite over NSW if any qualifying notice occurs. SA gets that edge because recent winter-style forecast tightness has been concentrated there, including the June 2026 forecast LOR2 and the very large number of SA forecast LOR2/LOR3 notices in Q3 2025. Offsetting that, South Australia is not in as weak a position as the unadjusted 2025 ESOO might imply: that ESOO flagged a 2026-27 SA reliability gap partly because Torrens Island B was assumed to retire on 30 June 2026, but subsequent official material says three Torrens B units (about 600 MW) were extended to 30 June 2028 while one unit still closes on 30 June 2026. Stage 1 of Project EnergyConnect has also already been operating with 150 MW transfer capacity since April 2025, though Stage 2 is not due until later in 2026. So SA is risky enough to lead the non-None states, but not by a wide margin. (aemo.com.au)
NSW is a very close second because it has converted tight conditions into actual LOR2 more often recently. AEMO’s quarterly reports show NSW had both actual LOR2 notices in Q4 2024, one of the three actual LOR2 notices in Q1 2025, and both actual LOR2 notices in Q4 2025. That recent conversion history matters. The reason I still keep NSW just behind SA is that the 2025 ESOO said NSW reliability risks remain low until much later in the horizon, and the immediate weather setup for early July is not especially threatening. (aemo.com.au)
QLD is third. The 2025 ESOO identified a small Queensland reliability gap in 2025-26, and AEMO notices show recent QLD forecast LOR2/LOR1 activity. But the same ESOO says Queensland reliability risks remain low across the horizon outside that near-term issue, and Brisbane’s early-July weather is mild. So I assign QLD a meaningful but still small share. Victoria is lower: the ESOO says 2026-27 reliability risks are lower there thanks to lower demand and more storage/VRE, although I leave a non-trivial tail because winter wind/network events can create reserve stress. Tasmania is lowest non-zero: the ESOO says 2026-27 Tasmanian risks are slightly higher because of planned winter Poatina unavailability, but still well below the reliability standard. (aemo.com.au)
Bottom line: my central estimate is that there is about a 14.4% chance of any qualifying actual LOR2-or-higher reserve notice in the window, and an 85.6% chance that no such notice occurs. Conditional on an event happening, I see SA and NSW as the two most likely regions, with SA only a narrow leader because its current winter reserve-tightness signals are stronger, while NSW has the stronger recent history of actual conversion. (aemo.com.au)
My base rate is that actual LOR2-or-higher notices are rare in the winter part of the NEM. In each of the last three Jul-Sep quarters that AEMO has reported separately (Q3 2023, Q3 2024, and Q3 2025), AEMO recorded zero actual LOR2 and zero actual LOR3 conditions, even though those quarters still had many forecast LOR2 notices. In Q3 2025 alone there were 27 forecast LOR2 notices and four forecast LOR3 notices, with South Australia accounting for 26 forecast LOR2 and all four forecast LOR3. (aemo.com.au)
That low winter base rate is reinforced by current official outlooks. AEMO’s 15 May 2026 winter outlook said the NEM had a stable winter outlook, with warmer-than-average conditions expected, more generation and battery capacity than last winter, similar coal and gas availability, and healthy southern gas storage. Separately, AEMO’s October 2025 ESOO update said the previously forecast 2026-27 reliability gap in South Australia was no longer expected after the delayed retirement of Torrens Island B. (aemo.com.au)
The main reason SA is still my top region is the live notice picture right now. In the current AEMO Market Notices feed retrieved on July 1, 2026, AEMO has an active forecast LOR2 for South Australia on 3 July 2026. The notice sequence starts with a 07:30-10:00 forecast LOR2 with reserve requirement 768 MW versus 625 MW available, and then updates it multiple times; one update widens the morning window to 07:30-11:00, and another briefly adds a small 14:30-15:00 period. AEMO says it is seeking a market response and has not yet estimated the latest intervention time. Because the resolution window starts at 00:00 AEST on 2 July 2026, this SA event is the most important near-term candidate for being first if any actual LOR2+ occurs. (aemo.com.au)
But I do not want to overreact to that single signal, because winter SA forecast LOR2 and LOR3 notices have often been canceled before becoming actual. In Q2 2025, 17 forecast LOR2 and three forecast LOR3 notices produced zero actual LOR2 or LOR3. In Q3 2025, 27 forecast LOR2 and four forecast LOR3 again produced zero actual LOR2 or LOR3. The early-July 2025 episode is especially relevant: SA had forecast LOR2 and LOR3 notices on 1-2 July 2025, and on 9 July 2025 operational demand reached 2,594 MW, only 34 MW below the winter maximum operational-demand record, yet there was still no actual LOR2 or LOR3. AEMO says many forecast notices did not eventuate because extra generation became available, net imports changed, demand changed, or network outages were altered. (aemo.com.au)
For the non-SA regions, NSW is second because recent actual LOR2 history is NSW-heavy: NSW had one actual LOR2 in Q1 2025 and two actual LOR2 notices in Q4 2025, while VIC and TAS had none in the reported quarters I reviewed. Still, AEMO’s 2026 ISP says winter reliability risk is shifting toward the southern mainland NEM, while Queensland’s reliability risk remains more summer-peak driven, which keeps QLD low and gives VIC a small but non-zero winter tail risk. (aemo.com.au)
Putting that together, my median view is still that no qualifying actual LOR2-or-higher reserve notice is issued between 2 July 2026 and 11 August 2026. SA gets by far the largest non-None share because it already has an active 3 July forecast LOR2 and would likely win on timing if that event turns into an actual notice. NSW is next on broader historical actual frequency, while VIC, QLD, and TAS remain only small tails. (aemo.com.au)
I treated this as a two-step problem: first estimate the probability that there is any qualifying actual LOR2-or-higher reserve notice between 00:00 AEST 2 July 2026 and 23:59 AEST 11 August 2026, then distribute that event probability across regions. AEMO declares an LOR when reserves fall below trigger levels, and its procedures and quarterly reviews show that many forecast LOR notices never become actual because extra generation becomes available or conditions improve. That forecast-to-actual gap is central here. (aemo.com.au)
My base rate is low. In Q3 2023, Q3 2024, and Q3 2025, AEMO reported zero actual LOR2 and zero actual LOR3 across the NEM. Q2 2024 and Q2 2025 also had zero actual LOR2/3 despite many forecast LOR2 notices. The only recent non-crisis Q2/Q3 actual LOR2 I found was one NSW event in Q2 2023. That makes None the modal outcome by a wide margin. (aemo.com.au)
Current 2026 conditions mostly push the probability downward. AEMO’s 15 May 2026 operational outlook described a stable winter outlook, warmer-than-average conditions, more than 3,000 MW of new generation and battery capacity added since last winter, similar coal and gas availability to last winter, and healthy fuel reserves. AEMO’s winter outlook material also said loss of load probability is generally low in NSW, TAS, SA, and VIC, and negligible in QLD, although hydro energy availability in the southern NEM warrants monitoring. (aemo.com.au)
For regional ranking, SA comes out first. The strongest live signal I found is that SA already had a June 2026 forecast LOR2 notice for 10 June, later followed by a forecast LOR1 for 17 June. Recent winter-quarter data also show SA repeatedly leading forecast scarcity counts, including 26 forecast LOR2 and 4 forecast LOR3 notices in Q3 2025, plus 8 forecast LOR2 and 3 forecast LOR3 in Q2 2025. At the same time, the October 2025 ESOO update removed SA’s formal 2026-27 reliability gap after the Torrens Island B retirement was delayed, so I do not read SA as broadly fragile; I read it as the most likely region if a winter reserve shortfall does break through to actual LOR2. (aemo.com.au)
NSW is my second choice. NSW is the only region with a recent actual LOR2 in the recent non-crisis shoulder/winter data set, via the Q2 2023 report, and AEMO’s Q4 2025 report also logged two actual LOR2 notices in NSW. Winter 2026 also includes several planned NSW transmission outages. I keep NSW close behind SA, but below it because the near-term weather is warmer in Sydney than in the southern mainland, and because SA has the fresher 2026 reserve-warning signal. (aemo.com.au)
VIC is third rather than second because recent actual LOR2 evidence is weak, but I still give it a meaningful tail. AEMO says Victoria’s winter demand sensitivity is increasing due to electrification, there are planned winter Victorian high-impact outages, and the immediate early-July weather around Melbourne is cool, windy, and under a severe-weather warning. (aemo.com.au)
QLD is lower because AEMO explicitly says winter LOLP is near zero there and Brisbane’s next 10 days are warm rather than stressfully cold. I still leave a small tail for surprise outages and for stressed-system outliers: AEMO’s report on 5 July 2022 shows Queensland did experience actual and forecast LOR conditions in a much more stressed winter regime. TAS is lowest because AEMO describes winter LOLP there as generally low, there are no planned Tasmanian high-impact outages for winter 2026, and recent Tasmanian LOR history is sparse. (aemo.com.au)
Numerically, my blended base-rate and current-conditions model puts the chance of any qualifying actual LOR2-or-higher notice in the window at about 15.1%, leaving 84.9% on None. I then allocate that event probability mostly to SA and NSW, with a smaller VIC tail and very small QLD and TAS tails. That yields: SA 6.2%, NSW 4.7%, VIC 2.0%, QLD 1.3%, TAS 0.9%, None 84.9%. (aemo.com.au)
Resolution is based on the first actual (not forecast) RESERVE NOTICE at LOR2 or LOR3 issued from 00:00 AEST on 2 July 2026 through 23:59 AEST on 11 August 2026. I make None the clear favorite because recent winter and shoulder-season history is strikingly sparse for actual LOR2-or-higher events. In Q3 2023, Q3 2024, and Q3 2025, AEMO recorded zero actual LOR2 despite 15, 5, and 27 forecast LOR2 declarations respectively. Q2 2024 and Q2 2025 also had zero actual LOR2, and Q1 2026 had zero actual LOR2 as well. By contrast, the recent actual LOR2 events were concentrated in warmer periods: Q4 2024 had two actual LOR2 events, both in NSW; Q1 2025 had three actual LOR2 events (NSW 1, QLD 1, SA 1); and Q4 2025 had two actual LOR2 events, both in NSW. That recent pattern argues for strong seasonality, with winter actual LOR2+ remaining unusual. (aemo.com.au)
Current forward-looking conditions also lean toward a quiet result. AEMO’s 15 May 2026 winter outlook says the NEM has a stable winter outlook, with warmer-than-average conditions expected, more than 3,000 MW of new generation and storage added since last winter, similar existing coal and gas availability, and healthy fuel reserves. The detailed winter readiness material says cold extremes are less likely than average; cold snaps can still occur and can push demand higher, particularly in NSW; LOLP is generally low across NSW, TAS, SA, and VIC and negligible in QLD; planned high-impact outages are lower in QLD, VIC, and SA than last winter; and there is additional battery capacity in every NEM region except Tasmania. The same material does note a dry-weather risk for southern hydro availability, so I do not treat the risk as zero. (aemo.com.au)
Conditional on there being any actual LOR2+, I rank SA first and NSW second. The main reason for SA on top is that recent winter LOR2 forecasting has overwhelmingly clustered there: in Q3 2023 SA had 15 forecast LOR2 notices and NSW had none; in Q3 2024 SA had 4 versus NSW 1; and in Q3 2025 SA had 26 versus NSW 1, while QLD had none in all three quarters and VIC/TAS had none in the latter two. That makes SA the most plausible winter region if a forecast shortfall fails to resolve. NSW is next because recent actual LOR2 events have mostly been NSW in warmer seasons, and AEMO specifically flags NSW as the place where winter cold snaps can still lift demand sharply. VIC is a smaller third-place possibility from southern-system dryness and outage interactions, while QLD is held down by AEMO’s ‘negligible LOLP’ language despite maintenance, and TAS remains lowest because planned HIOs are absent and historical LOR activity has been light. (aemo.com.au)
My quantitative base rate is low. Using the four most comparable recent quarters (Q2-Q3 2024 and Q2-Q3 2025), all of which had zero actual LOR2, a simple Gamma-Poisson estimate gives only about a 5-6% chance of at least one actual LOR2 in a 41-day window like 2 July to 11 August. I then shade that upward modestly rather than downward because AEMO still issued 2026 forecast LOR2 notices in SA in June and QLD in May, reminding us that sudden outages or interconnector issues can still create a first actual notice even in an otherwise benign winter. That leaves me at 93.0% for None, with the remaining 7.0% spread mostly across SA and NSW. (aemo.com.au)