Ask a planning team how many tools it takes to answer, "what should we build, and will it keep the lights on," and you'll get an uncomfortable number.
Ten tools.
Ten teams.
Ten sets of assumptions (that never quite agree with each other).
Somebody's resource plan says one thing, somebody's gas portfolio view says another, and nobody can fully explain why.
That gap is the reason the industry is rethinking how it models the grid, and it’s the problem PLEXOS® sets out to solve.
Modeling teams are spending too much time wrangling data and not enough time driving decisions. Part of that comes down to talent. Roughly 400,000 US oil and gas workers are approaching retirement, according to McKinsey, and the competition for technical expertise is only getting tighter as the grid gets more complicated. Meanwhile, the system itself has changed under everyone. Renewables have added volatility, congestion has gone from occasional to structural, and price formation now has to account for carbon policy and cross border flows that simply weren't part of the equation a decade ago.
When ten disconnected tools are trying to keep up with that, something breaks – and usually it is the trust in the numbers.
PLEXOS® takes a different approach: one platform and one engine, covering power, gas, hydrogen, heat, and transport, from real time operations all the way out to 20-year planning horizons.
Here's the part worth sitting with. Capacity expansion and resource adequacy aren't treated as separate products. They're the same model asking two different questions. One tells you what to build. The other tells you whether what you built will hold up.
Capacity expansion is about answering the big, expensive questions: what to build, when to build it, what technology to use, and when to retire what's already there. These are long horizon decisions, typically run out 20 to 50 years, and the objective is straightforward to state even if it's hard to solve. Minimize the net present value of total system cost, capital cost and production cost combined, formulated as a mixed integer problem. That last part matters. A resource mix that looks cheap to build but expensive to run isn't actually the lowest cost answer, and a model that only optimizes for capital cost will miss that every time.
PLEXOS® handles this with a level of detail that older approaches tend to gloss over:
Once you know what you're building, the next question is whether it actually keeps the system reliable. A lot of the future is uncertain, load, fuel prices, weather, regulations, demographics, and deterministic scenarios can only tell you so much about that uncertainty. This is where Monte Carlo simulation comes in.
PLEXOS® runs a sequential chronological commitment and dispatch across a wide range of historical weather years and random outages, layering in additional variables where needed, including transmission constraints, fuel availability, correlated outages, and the kind of extreme events that used to be rare and increasingly aren't. The output is a set of reliability metrics, like loss of load hours, loss of load expectation, and expected unserved energy, measured against the industry benchmark of one day in ten years.
What that looks like in practice is a real difference. In one example, a system started with a loss of load expectation of 1.79 and 149 GWh of expected unserved energy under its original resource mix. Once additional firm capacity was added and modeled through PLEXOS, those numbers dropped to a loss of load expectation of 0.115 and just 1.16 GWh of expected unserved energy, landing right at the one day in ten years standard. That's the whole point of running the stress test. It turns "this should probably work" into a number you can defend.
Here's the catch that trips up a lot of planners. A portfolio can look great on paper, least cost, and still fall apart once you test it probabilistically.
Because capacity expansion and resource adequacy run on the same engine and the same dataset in PLEXOS, that testing loop doesn't mean rebuilding the model or reconciling two sets of assumptions every time something doesn't check out. You expand, you test, and if it comes up short, you feed that back in and expand again.
A good example of this in action is Midcontinent Independent System Operator (MISO), the largest ISO in North America, spanning 15 states and Manitoba. In a recent webinar with Energy Exemplar, RaeLynn Asah, Senior Manager, Policy and Regulatory Planning, discussed how MISO recently ran its full Futures Process in PLEXOS® at scale for the first time, having previously used it mainly for regional assessments.
The four scenarios show how wide a planning range can get. Load growth alone spans from a 0.35% compound annual growth rate in Future 1 up to 2.1% in Future 3, with emissions targets ranging from a 40% reduction from 2005 levels to as high as 80%, depending on the scenario. By 2045, that spread shows up directly in installed capacity: Future 1 lands at 413 GW, Future 2 at 437 GW, Future 3 at 455 GW, and Future 4, the supply constrained scenario, at 501 GW.
The results were notable beyond the numbers too. Working from these four futures, the model justified building small modular reactors for the first time ever. Every scenario, including the one built around current supply chain constraints, came out resource and energy adequate, holding to MISO's reliability criterion of 0.1 day per year annually and 0.01 day per year seasonally.
It's also worth noting how this connects to the rest of MISO's planning pipeline. Member plans flow in through MISO's Generation Interconnection Queue and Expedited Queue, generation planning runs through PLEXOS, and the resulting plan then feeds MISO's downstream transmission planning process.
It’s worth being honest about too: the transition took a bit longer than usual, about a year and a half instead of the typical year, partly because it was a new platform and partly because of unusually different regulatory assumptions this cycle. MISO expects that timeline to shrink back down as the process matures.
PLEXOS® is used by transmission and system operators globally, and that matters beyond just methodology. Regulatory and stakeholder processes are adversarial by design, so when your analysis runs on a platform that operators across six continents rely on for their own reliability obligations, you're defending a methodology instead of a black box.
That's really the whole pitch in one sentence. Better modeling isn't about adding more tools. It's about giving planning teams one framework that can keep up with how complicated the grid has become, and MISO's experience is a solid proof point of what that looks like in practice.
Want the details behind the build-outs? Watch Dr. Joe Nyangon and Jonathan Surls of Energy Exemplar join RaeLynn Asah of MISO to unpack the Series 2 Futures, from load growth scenarios to how MISO tests every portfolio for resource adequacy. Stick around for the Q&A, where the audience didn't hold back. Watch the recording.