.png)
Overcoming the "Nervous" Tenant: DataBank’s Strategy for De-Risking Automated HVAC Optimization
DataBank is tackling the primary barrier to data center HVAC control sequence optimization—tenant "nervousness"—by deploying digital twin software as a validation sandbox. Despite hefty efficiency and sustainability goals, Databank faces a recurring hurdle: customers fear that AI-driven or automated BMS sequences might compromise critical uptime.
“We can only do a certain amount of controls until tenants get nervous,” said Jenny Gerson, who runs sustainability for DataBank, curing her presentation at NexusCon '25. “They don’t want everything automated. They don’t want AI running the BMS; there’s still a lot of manpower on the ground”.
To bridge this trust gap, DataBank implements a digital twin model that creates a high-fidelity replica of the facility to test sequences and optimization recommendations before applying them to live systems. This allows the engineering team to validate aggressive strategies in a remote environment, effectively turning the digital twin into a risk-mitigation tool.
DataBank operates on a co-location model where tenants pay a "PUE multiplier" to cover overhead. If the facility operates more efficiently than the target set in the Service Level Agreement (SLA), the owner captures the difference. “In our SLA, we’ll often have a certain PUE that we have to stick to... if we do better than that PUE, that's money in our pocket,” Gerson noted. A direct financial incentive.
By using digital twins to "take some risk out of the system," DataBank aims to push efficiency beyond basic LED retrofits and airflow optimization, especially in aging facilities that are 20 to 30 years old.
Learn more:
- Watch the full presentation from NexusCon 2025
- Sign up for the Nexus Labs newsletter to get five similar stories for owners each Wednesday:
DataBank is tackling the primary barrier to data center HVAC control sequence optimization—tenant "nervousness"—by deploying digital twin software as a validation sandbox. Despite hefty efficiency and sustainability goals, Databank faces a recurring hurdle: customers fear that AI-driven or automated BMS sequences might compromise critical uptime.
“We can only do a certain amount of controls until tenants get nervous,” said Jenny Gerson, who runs sustainability for DataBank, curing her presentation at NexusCon '25. “They don’t want everything automated. They don’t want AI running the BMS; there’s still a lot of manpower on the ground”.
To bridge this trust gap, DataBank implements a digital twin model that creates a high-fidelity replica of the facility to test sequences and optimization recommendations before applying them to live systems. This allows the engineering team to validate aggressive strategies in a remote environment, effectively turning the digital twin into a risk-mitigation tool.
DataBank operates on a co-location model where tenants pay a "PUE multiplier" to cover overhead. If the facility operates more efficiently than the target set in the Service Level Agreement (SLA), the owner captures the difference. “In our SLA, we’ll often have a certain PUE that we have to stick to... if we do better than that PUE, that's money in our pocket,” Gerson noted. A direct financial incentive.
By using digital twins to "take some risk out of the system," DataBank aims to push efficiency beyond basic LED retrofits and airflow optimization, especially in aging facilities that are 20 to 30 years old.
Learn more:
- Watch the full presentation from NexusCon 2025
- Sign up for the Nexus Labs newsletter to get five similar stories for owners each Wednesday:
This is a great piece!
I agree.