Midwest Office Complex Improves on Near-Perfect Energy Efficiency Rating

Multi-use facility uses AI to further optimize its 99 ENERGY STAR score.

Learn more

$25000

Cost Savings

780 tons

CO2e Eliminated

Overview

This site is a 2.2 million-square foot complex located in the heart of the midwest. As a for-profit facility, it is a mixed use of office and medical space. It also houses three large auditoriums, a fitness center, cafeteria, and other amenities for tenants.

‍

THE Challenge

Optimizing an already world-class facility

The facility is already doing an excellent job in their energy efficiency efforts. With a 99 ENERGY STAR score, it has a near perfect rating (scores range from 1-100). The ENERGY STAR score assesses a building’s physical assets, operations, and occupant behavior to provide a comprehensive snapshot of energy performance. To arrive at a score, the facility is compared to thousands of similar buildings across the United States, where the median score is 50.

Also aiding in its efficiency, the facility’s energy managers take advantage of 2,800 free cooling hours per year, on average. Free cooling is a process that helps to reduce the operating costs of HVAC systems. It occurs when buildings can leverage favorable outdoor air temperatures to chill water in the cooling system, bypassing the need for mechanical chillers powered by electricity.

In short, the facility is a world-class operation and sterling example of energy efficiency. Nevertheless, Tagup challenged their team to prove our artificial intelligence could further improve their operation — specifically surrounding the cooling system.

‍

LOCATION

Midwest, U.S.

CONTROL SYSTEM

Tridium Niagara N4

TOTAL SQUARE FOOTAGE

2,200,000

HVAC EQUIPMENT

Trane Centrifugal Chillers, BAC Towers

THE SOLUTION

A lightweight software implementation

Rightfully, the facility’s team did not immediately see the need for adding a cooling optimization software to their portfolio. However, after discussing possible savings and offering a risk-free deployment of Beacon, Tagup and facility representatives set out to gather necessary information. Factors like current control strategies, weather data, and utility rates were captured, as was an understanding of site implementation requirements.

Integrating Beacon into the facility’s building automation system was a turnkey process requiring no additional hardware and zero downtime. From there it was only a matter of letting Beacon aggregate data while the cooling system ran. From this point forward, the artificial intelligence software learns from system behavior to automatically improve its operational scheme.

‍

THE SUCCESS

Continuous learning and optimization

The facility had historically used a fixed temperature setting for the water supply entering the chiller plant. This fixed setpoint approach is often the go-to for energy and facility management teams because it typically ensures reasonable performance in all conditions. However, this kind of control logic is rarely optimized for efficiency. It doesn’t take into account the complex interactions between components resulting from external factors that affect cooling system performance, like weather conditions and cooling load variations based on tenant behavior. This can lead to energy waste because setpoints aren’t dynamically adjusted in real-time.

Beacon’s machine learning immediately operationalized optimal setpoints to minimize chiller plant operating costs. By using artificial intelligence to interpret available data and automatically adjusting setpoint controls, Beacon has ensured that the facility’s cooling system will operate at peak performance levels without the need for human intervention.

Beacon will save the facility $25,000, or about 5% of cooling operating costs in year one. It will see an emissions reduction equivalent to 780 tons of carbon dioxide. For a building as energy efficient as this facility, these cost and carbon emissions savings are substantial in value.

Cost savings and emissions reductions will continue to increase over time as the machine learning software more accurately trains its models.

‍