Sunairio

Attending the Gulf Coast Power Association (GCPA) Annual Fall Conference & Exhibition this week? So are we! The Sunairio team — including Director of Strategic Growth John C. Enny and Growth Associate Mason Shambeau — is on-site as a proud Gold Sponsor of this year's event. And we’re ready to tackle today’s most important power sector conversations… In today’s power market, extreme weather and grid volatility are no longer edge cases — they are the primary drivers of capacity risks, price spikes, and operational uncertainty. Whether you're looking to enhance electricity generation and demand forecasting, optimize asset value, or strengthen grid reliability risk management, we’d love to connect. If you’re in Austin for GCPA, drop by, meet the team, or send us a message to set up a time to chat.

Did you know… Sunairio provides energy forecasts (load, wind generation, solar generation, and net load) derived from the public forecast weather models right alongside our proprietary Sunairio ONE ensemble. We do the heavy lifting of managing the pipelines and handling the proper hourly interpolation (non-trivial!) of the public models that publish in 6-hourly time steps. This way you can easily quantify the likelihood of a public forecast solution by comparing their output to a Sunairio ONE calibrated percentile level. See below for an example: the IFS Ensemble and the AIFS Ensemble both show extreme members that result in 132+ GW of PJM RTO load on September 21, which is beyond the hourly Sunairio P99 forecast level. Could those load levels be verified? Sure… there’s a 1/100 chance.  Simple, fast, and powerful insights to help you manage risk.

We’re headed to The North American Energy Markets Association’s 2026 Fall Conference! With extreme weather events and market volatility increasingly driving grid reliability and power price fluctuations, managing risk has never been more critical. 📍Omni PGA Frisco Resort & Spa, Frisco, TX 🗓️September 21–23 Connect with Sunairio CEO Rob Cirincione and Director of Strategic Growth John C. Enny on the ground to learn how our next-gen forecasting helps power market participants tackle risk and make data-driven decisions. Event details: https://lnkd.in/eY27Zvzu

We’re excited to announce the launch of ESR (energy storage resource, i.e. battery) charging forecasts in ERCOT. Why is this important? 🟣 Battery charging is like any other electrical load on the grid in terms of its contribution to grid stress and price formation 🟣 Battery charging schedules are much more complicated to predict accurately than traditional demand resources because they’re a function of net grid balances 🟣 There’s currently about 21 GW of batteries installed in ERCOT (that’s a lot!) 🟣 ERCOT ISO *REMOVED* ESR/battery charging demand from its official “load” time series late last year. If you’re not including ESR charging in your ERCOT demand forecasts now, you’re missing a significant contributor to grid stress. Yes, you read that correctly. What ERCOT now reports as load is an incomplete measure of total demand on the grid. Want to see the full picture? Check out the screenshot below that shows Sunairio’s P05-P50-P95 hourly ESR charging forecasts for the past few days vs. actuals reported by the ISO.

So it looks like there’s one more sneaky heat event next week in the East before we can close the books on summer (we’ll have more to say about this soon), and then every power trader’s gaze will turn to…OUTAGE SEASON!!! (cue dramatic trumpet fanfare) If you’re as excited about generation outage patterns as we are, we think you’ll like the first plots below that measure PJM’s 7-day outage forecast error by forecast lag (days out) and forecast month. The trends are pretty striking: PJM chronically under-predicts generation outages and the error gets worse as the forecast lag increases. This holds for every month of the year and both sub-region groups that PJM provides outage forecasts in. That’s why we create our own proprietary outage forecasts that significantly outperform PJM’s own data. Year-to-date, our next-day outage forecast bias has been 38% (600 MW) better than the ISO’s official versions. Back to the incoming heat. Removing one MW of offline generation can be as impactful to the grid as adding one MW of load. Now take a look at the screenshot below (second image) from Sunairio that plots PJM’s official outage forecast (dark purple) for Sep 1-2 (when the heat arrives in the east) against our P01-P99 range and the actual outage levels for the past week. Actual outage levels have been hovering around 20 GW offline, which is the middle of our forecast for next week, but PJM has an official outage forecast of only ~4.5 GW – likely much too low. If you’re taking PJM’s forecast as truth, you’re drastically underestimating the amount of generation that will be offline next week…and missing a significant amount of the grid risk.

We’re headed to NAEMA’s 2026 Fall Conference next month in Frisco, TX! This year marks the 30th anniversary of the event, bringing together leaders across the energy sector. Connect with Sunairio CEO Rob Cirincione and Director of Strategic Growth John C. Enny on the ground to learn how our next-gen forecasting helps power market participants tackle risk and make data-driven decisions. Event details and registration: https://lnkd.in/gb-YTVNj

Exciting news for power traders, risk managers, and energy analysts. We’re thrilled to announce our new data partnership with Grid Status! Navigating today’s power markets isn't getting any easier. Extreme weather events, rapid renewable integration, and surging load growth mean that managing grid risk requires stronger insights. To build winning commercial strategies and mitigate downside risk, you need two things above all: next-gen probabilistic forecasting and uncompromising ground truth. Through this partnership, Sunairio users can now seamlessly access official ISO data directly within our platform and API. What does this mean for your team? 🎯 Streamlined workflows: Evaluate, backtest, and validate trading strategies in a single ecosystem — no more stitching together disparate datasets or doubting historical price accuracy. 📊 Instant validation: Contextualize and benchmark Sunairio’s grid forecasts against official ISO real-time and historical data at scale. 📉 Actionable intelligence: Better navigate power market fundamentals and tail-risk exposure with confidence. As our CEO, Rob Cirincione, put it: “The importance of quality data to serve as ground truth really can’t be underestimated. Sunairio customers can now build end-to-end workflows that train, test, and validate strategies in a single platform that leverages a trusted ISO data provider. No more combining data from different sources, and no more worrying that the historical prices aren’t correct.” Ready to streamline your grid forecasting and backtesting workflows? Read the full press release here: https://lnkd.in/gv6y5x8q

Many of you in the Midwest and East Coast have noticed the recent wildfire smoke. It’s not just affecting air quality—it’s actively suppressing solar generation. The catch? Many weather models don't simulate wildfire smoke. They predict solar radiation using only cloud cover and generic aerosol data. As a result, standard energy models will forecast peak solar generation even when a site is completely choked by smoke.  At Sunairio, we forecast solar generation using GHI and leverage weather model intelligence that resolves wildfire smoke. Look at today's 12Z HRRR run below: ➡️ The Problem: Parts of the upper Midwest are relatively cloud-free, yet GHI (bottom right) is drastically lower than expected. The culprit? Smoke (top right). ➡️ The Nuance: The patterns you’d expect from total cloud coverage don't match reality because upper and lower-level clouds attenuate radiation differently. To get an accurate prediction of solar generation, you have to account for all of these moving parts simultaneously. If solar generation is what drives your business, irradiance is the variable you need to forecast. Relying on cloud cover alone is an oversimplification of a highly complex problem.

Early fireworks in PJM last week! In our latest blog post we characterize the July 1-3 grid event (including the emergencies on July 2-3) and discuss how well Sunairio did anticipating it. We cover: - Weather extremes as seen through Sunairio High-res Earth Data (image below) - The temperature-load relationship in PJM - Demand Response activations and observed load reductions - Two grid emergency (NERC EEA 2) calls - The accuracy of Sunairio’s forecasts – including hub prices Read on! https://lnkd.in/gcTwbyrN

Just how hot is it forecasted to be next week in PJM? Extremely hot – throughout the RTO. Our median population-weighted forecast for PJM Interconnection for Thursday Jul 2 is 96 deg F. Here are the all-time records according to Sunairio High-resolution Earth Data (data back to 1950): 🏆: 97.8 (Jul 07, 2012) #2: 97.1 (Sep 02, 1953) #3: 96.8 (Jul 21, 2011) #4: 96.5 (Aug 31, 1953) #5: 96.5 (Jul 06, 2012) Why don’t these look more extreme? PJM is huge. This is a weighted-average hourly temperature that includes places like Richmond and DC, but also Chicago and Cleveland. It’s really unusual to have upper-90s temps realize over this entire footprint – at the same time.