PCI Energy Solutions finds the money hiding in the power bill

An electricity bill is an unlikely place to look for a good business story. It has numbers, abbreviations and the general charm of a tax form. Yet Guzman Energy found something rather more attractive in its settlement data: a billing error worth more than $1.2 million.

PCI Energy Solutions’ customer account says Guzman recovered the money using shadow settlement and validation tools. Shadow settlement means calculating what a market bill should contain, then comparing that calculation with the actual statement. The company had been billed incorrectly in the Western Energy Imbalance Service market. A shared system made the discrepancy visible.

The invoice with a seven-figure mistake

The revealing detail is where the value appeared. Guzman did not need a new generating station to recover that money. It needed a better view of transactions already taking place. PCI’s back-office tools ingest statements, reconcile prices and volumes, and retain records for disputes. A trading decision can be brilliant and still lose money when its final bill is wrong.

“That alone paid for the project.”Christin Hanley · Director of Data, Guzman Energy

Hanley’s assessment turns an abstract software purchase into a concrete result. It also leaves room for sensible restraint: one recovered error is a customer outcome, not a guaranteed return for every buyer. The lesson is narrower, and more useful. If the organization cannot independently reconstruct what it owes, it has outsourced part of its judgment to whoever sends the invoice.

An algorithm meets an incentive

PCI began with two University of Oklahoma professors, Fred Lee and Art Breipohl, and the problem of scheduling generation. Which units should operate, and when? Their company was formed in 1992. But Breipohl’s recollections contain a delicious commercial inconvenience: utilities could believe the software saved fuel while seeing little reason to buy it. Fuel expenses passed through to customers did not necessarily change utility profit.

Competition altered that calculation. Lower costs acquired a different commercial meaning when electricity could be traded. The early difficulty was the incentive around the product. The founders also discovered that teaching could lead to sales: an industry short course brought them into contact with future customers.

A megawatt has paperwork

Today PCI occupies the stretch between physical energy assets and their financial consequences. A utility forecasts demand, schedules generation, buys fuel, submits market offers, monitors exposure and checks settlement statements. Each step changes the information the next team needs. Passing it around by hand creates opportunities for delay and disagreement.

PCI ETRM covers energy trading and risk management across power, gas and other commodities. Front-office tools capture deals and support scheduling. Middle-office tools monitor positions, valuation, limits and approvals. Back-office tools reconcile transactions and produce settlement and accounting outputs. The attraction is a record that remains useful as the transaction moves through the organization.

GenManager handles bid-to-bill workflows. GenTrader addresses portfolio optimization and generation planning. Transmission products manage scheduling and related accounting; outage tools coordinate when generation or transmission equipment will be unavailable. The expertise is unusually specific. A power trader’s schedule and a plant operator’s maintenance plan must reflect the same practical limits, even though the two people have different jobs.

Forecasting adds another layer. PCI Forecaster uses machine learning and analytics for demand, wind, solar and locational marginal prices. It can combine forecast sources, refresh data and feed bidding workflows. A prediction becomes more useful when it reaches the person making the next decision without a detour through someone else’s spreadsheet.

The customers include investor-owned utilities, public power organizations, cooperatives, independent producers and traders. Their needs differ. Holy Cross Energy selected e-Tag+ in September 2026 for visibility and reporting around power scheduling tags. The tool can stand alone or join a wider platform. Buying PCI need not mean buying every part of PCI.

The trouble with a system nobody dares touch

At Ørsted Americas, the trigger was a legacy ETRM system that had accumulated enough customization to make change slow and risky. An AWS account of the project describes an expanding U.S. onshore portfolio exceeding 5 GW. The old system was becoming an obstacle to the next product, rather than a dependable foundation for it.

Ørsted evaluated vendors and chose PCI’s cloud ETRM on AWS, citing design, automation and API capabilities. A small joint team paced the implementation. That detail matters: moving to cloud software still requires people to agree on records, connections and operating procedures. The infrastructure can change more quickly than the habits around it.

“We knew we had to reset.”Soren Hansen · Head of Asset and Trading Risk, Ørsted Americas

Consumers Energy faced a different mismatch. PCI’s battery case study describes operating a 100 MW battery with approaches inherited from thermal generation and pumped hydro. BatteryTrader, integrated with Forecaster, connected forecasting and real-time offers. PCI reports offer construction taking ten minutes and real-time returns 30% above day-ahead results in that case. The comparison concerns those trading approaches within the deployment; it is not a universal uplift from purchasing software.

Pattern Energy’s SunZia deployment supplies another example. PCI announced in July 2026 that forecasting and outage-management tools supported the project’s operational preparation. A renewable project needs to coordinate test energy, forecasts and equipment availability before commercial operation. The software has work to do before the first routine trading day arrives.

The bill for the bill-checker

PCI sells modular enterprise software, recurring SaaS access and the services needed to implement and support it. Its cloud relationship with AWS includes an Energy ISV Competency and Marketplace availability announced in 2024. Public materials put considerable weight on implementation support. The customer buys a continuing relationship with people who understand changing power-market requirements.

There are credible alternatives. ION’s Openlink serves complex multi-commodity trading, risk and physical logistics; Allegro also addresses utilities and energy companies. Many buyers additionally compare a platform with their own spreadsheets, homegrown applications and specialist tools. PCI’s positioning centers on linking asset optimization with market operations and accounting. An integrated platform is attractive when the connections themselves are expensive to maintain.

The business now has a new prospective owner. Mitsubishi Electric agreed on August 20, 2026 to acquire all PCI equity for a base price of $1.4 billion, subject to closing adjustments. Its announcement envisages combining PCI’s optimization expertise with control technology and broader sales channels. PCI says the transaction requires approvals and closing conditions, and that the companies will operate separately until completion.

Copy the question before the software

For a buyer, the useful starting point is one troublesome workflow. Reconstruct a market bill. Trace a trade into its schedule. Compare a forecast with what the bidding desk actually received. Find the point where someone copies, repairs or distrusts the data. Then ask vendors to demonstrate that journey with the organization’s own market rules and asset constraints.

That exercise also exposes the conditions for success. A shared platform needs reliable inputs, working integrations and people responsible for maintaining the model. Forecasts remain uncertain. Batteries have physical limits. A small operator with straightforward transactions may have less to gain from a broad enterprise deployment than a utility coordinating several markets and asset types. The test is whether the recovered time and value justify the complexity of implementation.

PCI’s history keeps returning to that practical arithmetic. The professor’s algorithm needs an incentive. The forecast needs a route to the trading desk. The trade needs a bill someone can verify. Guzman’s recovery gives the argument its most memorable number. Sometimes the money is already in the business. Someone simply has to check where it went.