From Three Weeks to Three Days: AEP Speeds IRP Updates with PLEXOS® Intelligence

American Electric Power’s (AEP) seven-person Integrated Resource Plan (IRP) modeling team used PLEXOS® Intelligence to cut a three-week data update into a day, on the same complex workbook that had defeated an earlier attempt at coded automation.

The challenge: rising demand

AEP serves 5.6 million customers across 11 states, runs the nation’s largest electric transmission system, and operates roughly 33 GW of generating capacity. Planning for all of it falls to an IRP modeling team of five people.

That team’s workload is climbing fast. Load growth, much of it from data centers, is driving more Requests for Proposals (RFPs), more new-generation analysis, and more IRP cycles across AEP’s operating companies, each with its own regulators and deadlines.

Getting new candidate resources into PLEXOS® was one of the team’s biggest bottlenecks. Every resource carries financial and operational inputs, and those inputs change whenever RFP bids arrive, cost curves shift, or internal projections are revised. For emerging technologies, inputs can shift dramatically.

All of it lives in one large Excel resource screening workbook, fed by teams across AEP and by outside vendors. It holds tab after tab of build costs, operating costs, ratings, and time series, with adjustments for each operating company layered on top. Updating the model meant working through it tab by tab and entering the data by hand, a job that took three weeks every time.

AEP had already tried to code its way out. It didn’t work.

“We built automation for a past IRP cycle, but the workbook had so many nuances that we kept rewriting the code. Maintaining it took as much work as doing it by hand, so it was a wash. We hoped AI would be smart enough to handle those nuances.” – Mohamed Abukaram, Director, Resource Planning & Operational Analysis, AEP

That’s the real test for AI in energy modeling. Clean demo data is easy. A production workbook with years of nuance is not.

How AEP used PLEXOS Intelligence in its production workbook

Proof of concept in the first hour

AEP was among the first teams to use PLEXOS® Intelligence. Working on its own production workbook rather than a test dataset, the team reached proof of concept within the first hour and refined it over a few more. With regulatory deadlines looming, that speed mattered as much as the time saved.

Plain-language instructions, real data

Analysts described what they needed in plain language and pointed PLEXOS® Intelligence at the workbook. It read in the candidate generators, populated the model, and mapped detailed cost and operating parameters from across the workbook’s tabs.

Errors caught before they reach a run

PLEXOS® Intelligence also flagged what was missing: generators with no fuel defined and outage properties left blank. It spotted batteries set up as generators and checked with the team before correcting them. Issues like these hide easily in a large workbook. Here, they were resolved before any model ran.

Analysts stay in control

The team chose what to hand off, reviewed every change, and approved the inputs before anything moved forward. Edits were made on a copy of the study, and every edit stayed visible and traceable. That is the kind of record regulatory review expects.

A playbook that gets sharper

As the work progressed, the team saved what worked as reusable skills. These instructions tell PLEXOS® Intelligence which tabs hold which data and how to map it. The next update starts from a proven playbook, not from scratch, and it gets smarter each time.

Proven on the PLEXOS® engine

The inputs PLEXOS® Intelligence built went straight into a full simulation, which ran successfully.

“[PLEXOS® Intelligence] did 70 to 80 percent of the work, and we got a successful model run based on those inputs.” – Mohamed Abukaram, AEP

The results: faster IRP input updates

“[PLEXOS® Intelligence] is going to save so much time because we do this manually. Working on a resource screening workbook can take maybe two weeks. My guess is now it will take a day.”  – Papa Yaw Owusu-Obeng, Principal Resource Planning Analyst, AEP

What's next for AEP

AEP is already lining up its next workflows:

• Market and commodity inputs. Market prices, loads, and other hourly inputs are next in line for AI-led updates.
• Bid evaluation. The  AEP team routinely evaluates bid responses for new generation. Moving from a manual process to integrate the operational parameters of these responses into PLEXOS® Intelligence is expected to significantly improve quality and increase cadence.
• Dashboards for decision-markers. The team is building dashboards in PLEXOS® Intelligence to get results in front of leadership faster. 

Every workflow handed off means more IRP scenarios before regulatory deadlines, fewer input errors, and faster answers for internal and external stakeholders.

“[PLEXOS® Intelligence] keeps getting better by the day. Right now, we think of it as another member of the team.” – Mohamed Abukaram, Director, Resource Planning & Operational Analysis, AEP

The future of PLEXOS® is intelligent, and it’s here

PLEXOS® Intelligence is the PLEXOS® you trust, now fully intelligent: the energy decision platform that takes you from modeled data to your next move in minutes, not weeks. 

Analysts describe what they need in plain language. PLEXOS® Intelligence then prepares data, builds and edits the model, runs the simulations, and explains the results, all on the team’s own models. Analysts approve every change, and every scenario runs on the physics-based PLEXOS® engine that ISOs and regulators already rely on. The result: answers fast enough to act on and rigorous enough to defend.