Introduction
A leading global industrial gas manufacturer, with operations spanning dozens of countries and tens of thousands of employees worldwide, is a major producer of industrial gases serving critical sectors such as energy, chemicals, and manufacturing. With significant electricity demand across its production facilities, the company faced rising energy costs and mounting exposure to volatile electricity markets.
Challenges
In Texas, where several of its largest plants operate within the Electric Reliability Council of Texas (ERCOT) grid, the company incurs high electricity costs during high demand or tight supply periods—particularly in the summer months when exposed to Four Coincident Peak (4CP) transmission charges. Missing a 4CP event or reacting too late to a real-time price spike can result in hundreds of thousands of dollars in avoidable cost. Managing these risks without impacting production and safety is critical.
To address this, the company sought an advanced AI-driven solution capable of more accurately predicting 4CP events, enabling operational teams to manage consumption, reduce peak-related charges, and maintain production commitments. Prior to engaging C3 AI, the manufacturer had developed an in-house AI-based solution that predicted most—but not all—4CP events, leaving meaningful cost savings unrealized.
The Solution
The company partnered with C3 AI to implement C3 AI Energy Management on Microsoft Azure in just 26 weeks across three ERCOT facilities. The solution unified more than five years of historical grid and market data across 14 datasets and established live data pipelines refreshing every 15 minutes, creating a continuously updated foundation for forecasting and decision-making. Four ML models were trained to forecast system-wide load, predict 4CP events, and anticipate real-time price volatility. These capabilities provided a real-time, unified view of grid conditions and forecasts across sites.
Results
C3 AI Energy Management empowers energy managers to monitor conditions in real time, coordinate across sites, and take timely curtailment or load-shifting actions, unlocking up to $1.4M in additional annualized savings for the manufacturer. The C3 AI solution also achieved 100% recall of 4CP events, capturing peak events missed by the internal model. Given the initial success, the manufacturer is scaling the solution beyond ERCOT, beginning with PJM Interconnection, followed by Midcontinent Independent System Operator (MISO) and New York Independent System Operator (NYISO), as part of its multi-region energy optimization strategy.
Solution Architecture
Proven results in weeks, not years
Get insights into C3 AI’s capabilities, enterprise AI best practices, and highest-value use cases.
Gain insights into the C3 AI Platform's capabilities, its model-driven architecture, and test it against your company's sample data set.
Identify a high-impact business problem and collaborate with the C3 AI team to rapidly build an AI application that solves it.
Scale and deploy a tested C3 AI application into production. Incorporate user feedback and optimize algorithms to drive maximum economic value.
Download the PDF to read the full story — Download