How a Bitcoin Mining Facility Saved $1.5M in Energy Costs

Overview

A Bitcoin mining facility in Texas runs at 10 MW of continuous load, the kind of operational profile that makes energy cost management existential, not optional. In ERCOT, where real-time prices can spike from $30/MWh to over $2,000/MWh within minutes, mining operations face a binary choice: run through price events and absorb the cost, or curtail and lose hash rate. Neither option is sustainable without a structured approach to knowing when each decision actually makes financial sense.

The facility's energy spend was its largest recurring expense. In 2023, that meant managing a particularly volatile summer with no reliable way to predict when the four coincident peak (CP) events would occur.

‍

Challenges

Managing coincident peaks manually demanded significant time and operational attention. The team was monitoring markets themselves, making judgment calls with incomplete information, and inevitably catching some events late. At the same time, high real-time price events occurred frequently enough that avoiding them on instinct wasn't a repeatable strategy.

The facility also wasn't participating in demand response programs, meaning it was only on the cost side of the ledger when load reductions could have generated revenue.
‍

‍

"Every summer, coincident peaks cost us more than they had to. We were making the calls ourselves with no forecasting behind us, and we knew we were missing out on potential benefits."

Solution

Arcobi (Formerly Arcus Power) addressed the problem across three tracks simultaneously:

1. Price Event Avoidance via AI ForecastsThe facility set a customized curtailment threshold of $100/MWh. When AI Forecasts signaled that real-time prices were approaching or expected to breach that threshold, the facility curtailed mining load proactively. The threshold was calibrated against the facility's revenue and downtime tolerance, not a generic number.

2. Coincident Peak Prediction and Automated AlertsDataHub delivered automated ERCOT coincident peak alerts that flagged the four highest-load hours of the year — the events that set transmission charges for the following twelve months. In the summer of 2023, the facility successfully avoided all four CP events, eliminating what would have been a six-figure annual transmission cost increase.

3. Demand Response Program EnrollmentWith Arcobi's support, the facility enrolled in ERCOT demand response programs, converting its curtailment capacity from a cost-avoidance mechanism into a revenue source. Load reductions that previously happened at a net loss now generated payments.

‍

Results

Across price avoidance, coincident peak mitigation, and demand response participation, the facility saved $1,556,307 in its first full year with Arcobi.

Total operational downtime came to 547.5 hours — 6.25% of the year. The facility maintained an average load of 10 MW and avoided all four 2023 coincident peak events.

The more important shift: the facility stopped reacting to ERCOT and started running against a forecast. Energy management moved from a reactive overhead function to a proactive cost center with a clear financial return.

"We avoided every coincident peak last summer. For a facility our size, that's not a rounding error — that's over half a million dollars that stays in the business." — [COO, Texas Bitcoin Mining Facility]

‍

This case highlights the potential for energy-intensive industries like cryptocurrency mining to leverage advanced energy management solutions. By employing predictive analytics and tailored demand response strategies, businesses can transform the challenges of energy price volatility into opportunities for savings, resilience, and sustainability. Arcus Power’s approach not only minimized costs but also enhanced the company’s ability to operate efficiently in a complex energy landscape.