Podcast Summary: Currents (Norton Rose Fulbright)

Episode 333: Rethinking Forecasts in an Uncertain Climate
Host: Todd Alexander
Guest: Rob Sorencion, CEO of Scenario
Release Date: February 5, 2026

Overview of the Episode

In this thought-provoking episode, Todd Alexander sits down with Rob Sorencion, a veteran power and gas trader and now CEO of Scenario, to examine the escalating dangers facing power grids due to extreme weather events and the growing inadequacy of traditional weather forecasting models. Rob elucidates why current models are no longer fit for the evolving landscape and introduces Scenario 1, a new forecasting approach aimed at bridging the gap between weather and grid risk. The conversation delves into the technical, commercial, and behavioral drivers behind forecasting challenges, and Rob provides candid insights into industry incentives, extreme weather risk, and the future of renewable project forecasting.

Key Discussion Points & Insights

1. Rob Sorencion’s Background & Scenario’s Mission

• (00:43–01:48)
  • Rob’s career path: Engineer by training, extensive experience as a power and gas trader with Constellation, BETM/NRG, and Mitsubishi.
  • Founding Scenario due to the increasing difficulty in managing weather-driven risk as a trader:

    “Broadly, what we do at Scenario is generate high resolution, high fidelity, both historical and forecast data of weather, power generation, assets and our markets.” (01:27, Rob Sorencion)

2. The Critical Link Between Weather, Volatility, and Power Markets

• (01:54–02:37)
  • Almost every major reliability and economic challenge in power markets is weather-driven.
  • Traders spend enormous time interpreting weather models; minor changes can make major differences.

3. Unique Challenges in Translating Weather Data to Grid Risk

• (03:13–08:44)
  • Most weather data is public, but the value is in translation, not raw information.
  • Three main translation issues:
    • Temporal Resolution: Government models often only provide forecasts every 3–6 hours, but grid operators need hour-by-hour (or better) forecasts.
    • Spatial Resolution: Accurate wind/solar modeling requires much finer spatial resolution than typically provided (site-level vs. regional).
    • Extremes, Not Averages:

      “In ERCOT, the Texas power market, right, a third of the annual market value...accrues from less than 1% of the hours per year.” (07:11, Rob Sorencion)

    • Public models often fail to predict and model the rare, extreme events that drive grid risk and value.

4. Renewables, Storage & Peak Price Shifts

• (08:44–12:06)
  • Storage and additional gas–while helpful–do not eliminate volatility; they shift it.
  • The move to renewables (particularly solar) is pushing peak risk hours later into the evening.
  • Market impacts:

    “You’re moving what used to be a three, four hour...peak, and you’re squishing it into one or two hours.” (11:05, Rob Sorencion)

  • The possibility that existing price caps may not be sufficient to incentivize new investment due to compressed, intense risk hours.

5. The Pitfalls of Traditional Production Forecasts & Incentives

• (12:06–16:27)
  • Production forecasts (e.g., wind/solar P50s) are often too optimistic.
  • Incentive misalignment: Developers benefit from bullish forecasts, lenders rarely feel the downside, and the risk ultimately falls on equity owners.
  • Fragmentation:

    “We’ll talk to…development side [of IPPs], they’ll say...everything needs to be consistent...If we talk to the group that manages the fleet…they’ll say, oh, that’s great. I just want the most accurate projection of this…it’s almost like having two sets of books…” (15:23, Rob Sorencion)

6. Climate Change and Increasing Unpredictability

• (16:27–18:59)
  • The increase in weather unpredictability is real and significant.
  • The problem isn’t gradual changes in average conditions but the exponential rise in frequency of extreme events:

    “You in some cases exponentially increase the frequency of extremes…If you’re a utility, going from one problem day to 10 problem days, that’s a big deal.” (17:22, Rob Sorencion)

  • Existing planning often ignores these changes, creating blind spots in reliability assessments.

7. Scenario 1: A New Approach to Forecasting

• (18:59–22:48)
  • Scenario 1 is an “omniscale next generation ensemble”—generating thousands of forecast scenarios to capture uncertainty and extremes.
  • Ensemble approach (like Monte Carlo simulation):

    “We generate a thousand…[ensemble members] so that you can see the tail events in the extreme. That’s the ‘E’ [in Scenario 1].” (20:10, Rob Sorencion)

  • Hybrid model: Uses machine learning and statistics atop traditional data to cover timescales from hours to decades.
  • Designed to deliver a seamless, business-relevant forecasting platform that’s consistent across all horizons (hourly, monthly, decadal).

8. The Limits and Role of AI in Forecasting

• (22:48–24:54)
  • Scenario’s approach is more hybrid/statistical than pure AI, especially for modeling tail risk.
  • Key commercial limitation of black-box AI:

    “The AI is not going to do a great job in the extremes because you don’t see them…You can’t train on it.” (23:18, Rob Sorencion)

  • Customers need transparency and traceability—‘why’ as well as ‘what’—especially in high-stakes investment decisions.

9. Who Uses Advanced Forecasting & For What?

• (24:54–28:52)
  • Three primary audiences:
    • Traders: For operational, short-term market moves.
    • IPPs (Independent Power Producers): Primarily for managing/hedging risk of operating fleets.
      • Subset: Corporate energy buyers, managing portfolios of PPAs and seeking to minimize risk and optimize future contracts.
    • Reliability Organizations (Utilities/Grid Planners): For long-term planning, yet most still rely on outdated, climate-ignorant assumptions.

      “Still today...you’ll see...‘we assumed weather from 1990 to 2020 and...did not account for climate change.’ It’s in the document.” (27:15, Rob Sorencion)

  • Rob highlights a disconnect between planning and operational reality, and the urgent need to close this gap.

Notable Quotes & Memorable Moments

• On the urgency of extremes:

  “It’s the extremes that matter in the power sector, in power markets.” (06:10, Rob Sorencion)

• On market incentives:

  “Projects get built by people who are incentivized to build projects, and projects look better when they have higher production.” (14:08, Rob Sorencion)

• On planning for climate change:

  “I think that’s also kind of just amazing in the time we’re in...how critical...the whole planning study is to try to figure out can the grid withstand extreme events. And I think there’s low hanging fruit…” (27:50, Rob Sorencion)

Timestamps for Key Segments

| Segment | Topic | Timestamp | |---------|-------|-----------| | Introduction, Rob’s background | 00:29–01:48 | | Why trading demands sophisticated weather insights | 01:54–02:37 | | How weather data is (mis)translated to grid risk | 03:13–08:44 | | Renewables, storage, and price volatility | 08:44–12:06 | | Overly optimistic wind/solar forecasts & incentives | 12:06–16:27 | | Weather unpredictability & climate’s growing impact | 16:27–18:59 | | What is Scenario 1? | 18:59–22:48 | | Limits of AI for rare events & client needs | 22:48–24:54 | | Who uses the forecasts & how | 24:54–28:52 |

Conclusion

This episode unpacks the critical challenges of forecasting in a rapidly changing climate, with a candid look at misaligned incentives, technical limitations, and the urgent need to forecast not only what is likely, but what is possible. Rob Sorencion’s remarks offer not just a technical roadmap, but an industry call-to-action: Planners, traders, and investors can no longer rely on dated models in the face of volatility and extreme risks, and must evolve toward richer, more nuanced understanding—grounded as much in commercial reality as atmospheric science.