.webp)
The Boring Levers You Need to Start with for HVAC Sequence Optimization
When we think of HVAC Optimization, most immediately envision the cutting-edge opportunities. We’ve heard of Microsoft using machine learning to optimize on/off schedules and Amazon enabling AI to control their discharge air setpoints. These are real energy-saving opportunities that have only recently become feasible thanks to modern tech, but they are also being pursued by organizations that have been focused on optimization for decades.
For the average building owner trying to get a handle on their energy use, the boring optimization strategies win out.
Regardless of how simple or complex your optimization approach is, there are essentially only 3 levers you have to make optimization adjustments:
- Schedules - When equipment runs
- Setpoints - The target your equipment is aiming for
- Sequences - The control logic that defines what your equipment does
Let’s break down the simplest ways that building owners are tweaking these 3 levers to get the most energy savings.
Schedules
Gabe Sandoval of UCSF Health calls scheduling the biggest energy saver of his entire optimization program. The challenge is simple: most buildings aren’t used on a consistent, reliable schedule. An operator bumps a schedule up a couple of hours for a special event or because a building isn’t warming fast enough, and now the building runs two hours longer indefinitely. While schedule updates can be handled by advanced algorithms and occupancy metrics, has your organization started by asking: when was the last time we audited our schedules, and who is responsible for keeping them accurate?
Setpoints
Assuming the schedule is right, the next lever is the target itself, and the most common mistake is chasing one that's more extreme than the building needs. The highest-value setpoint adjustments are hiding in undramatic places, like widening the deadband between heating and cooling so equipment stops fighting itself. Rob Brimblecombe at Monash University clocked roughly 6% energy savings per degree of deadband he opened up. Before you point an algorithm at your setpoints, has anyone asked which targets are stricter than the building actually requires, and why?
Assuming your deadbands make sense, setpoint drift is often spearheaded by pesky overrides. Patrick Testoni of UC Santa Cruz referred to it as the “technological whack-a-mole”. He can’t stop other groups from setting overrides if they have access to the BAS, but he can set up rules to detect when that’s happening, so they can plan to release overrides effectively.
Sequences
The third lever is the control logic itself, and it’s the hardest of the three because it requires a fundamental understanding of the code and design intent. Last week, we talked about the hospital owner who was sure an HVAC energy audit would come up dry because Guideline 36 was already installed (spoiler alert — they were wrong, the building had drifted significantly).
Assuming your building was designed or retrocommissioned with suitable sequences like Guideline 36, the drift results from changes in personnel operating the building and changes in the building itself.
At Texas Tech, Brandon McCoy watched the programmers who built their sequences retire around 2017, and the skills never transferred, leaving "a lot of changes and band-aids" baked into the logic and no one in-house who could undo them. He’ll be sharing more of this story at NexusCast (June 17th, 2026).
Patrick Testoni of UC Santa Cruz describes the same rot from the other side: a new wing is added, the sequence is tuned for the new equipment, and the wire sheets are "revved and revved and revved" until someone has to start over.
Building owners are routinely capturing 10%-20% energy savings just by returning the building to how it was supposed to be run. Before you shop for an autonomous overlay, do you know who actually owns your control code, and whether anyone on your team can still read it?
The good news for building owners with immature optimization programs: the boring levers above are within reach, and will save you the most energy.
The good news for building owners with mature optimization programs: the rollercoaster of AI adoption is opening up increasingly reliable options for autonomous optimization through advanced supervisory control.
This article is part of our Building Owner Signal series. We're highlighting patterns we see across conversations with building owners. These short insights focus on operational risks, emerging priorities, and what owner teams should pay attention to now.
Register for the next Nexus Labs event.
Sign up for the newsletter to get 5 stories like this per week:
When we think of HVAC Optimization, most immediately envision the cutting-edge opportunities. We’ve heard of Microsoft using machine learning to optimize on/off schedules and Amazon enabling AI to control their discharge air setpoints. These are real energy-saving opportunities that have only recently become feasible thanks to modern tech, but they are also being pursued by organizations that have been focused on optimization for decades.
For the average building owner trying to get a handle on their energy use, the boring optimization strategies win out.
Regardless of how simple or complex your optimization approach is, there are essentially only 3 levers you have to make optimization adjustments:
- Schedules - When equipment runs
- Setpoints - The target your equipment is aiming for
- Sequences - The control logic that defines what your equipment does
Let’s break down the simplest ways that building owners are tweaking these 3 levers to get the most energy savings.
Schedules
Gabe Sandoval of UCSF Health calls scheduling the biggest energy saver of his entire optimization program. The challenge is simple: most buildings aren’t used on a consistent, reliable schedule. An operator bumps a schedule up a couple of hours for a special event or because a building isn’t warming fast enough, and now the building runs two hours longer indefinitely. While schedule updates can be handled by advanced algorithms and occupancy metrics, has your organization started by asking: when was the last time we audited our schedules, and who is responsible for keeping them accurate?
Setpoints
Assuming the schedule is right, the next lever is the target itself, and the most common mistake is chasing one that's more extreme than the building needs. The highest-value setpoint adjustments are hiding in undramatic places, like widening the deadband between heating and cooling so equipment stops fighting itself. Rob Brimblecombe at Monash University clocked roughly 6% energy savings per degree of deadband he opened up. Before you point an algorithm at your setpoints, has anyone asked which targets are stricter than the building actually requires, and why?
Assuming your deadbands make sense, setpoint drift is often spearheaded by pesky overrides. Patrick Testoni of UC Santa Cruz referred to it as the “technological whack-a-mole”. He can’t stop other groups from setting overrides if they have access to the BAS, but he can set up rules to detect when that’s happening, so they can plan to release overrides effectively.
Sequences
The third lever is the control logic itself, and it’s the hardest of the three because it requires a fundamental understanding of the code and design intent. Last week, we talked about the hospital owner who was sure an HVAC energy audit would come up dry because Guideline 36 was already installed (spoiler alert — they were wrong, the building had drifted significantly).
Assuming your building was designed or retrocommissioned with suitable sequences like Guideline 36, the drift results from changes in personnel operating the building and changes in the building itself.
At Texas Tech, Brandon McCoy watched the programmers who built their sequences retire around 2017, and the skills never transferred, leaving "a lot of changes and band-aids" baked into the logic and no one in-house who could undo them. He’ll be sharing more of this story at NexusCast (June 17th, 2026).
Patrick Testoni of UC Santa Cruz describes the same rot from the other side: a new wing is added, the sequence is tuned for the new equipment, and the wire sheets are "revved and revved and revved" until someone has to start over.
Building owners are routinely capturing 10%-20% energy savings just by returning the building to how it was supposed to be run. Before you shop for an autonomous overlay, do you know who actually owns your control code, and whether anyone on your team can still read it?
The good news for building owners with immature optimization programs: the boring levers above are within reach, and will save you the most energy.
The good news for building owners with mature optimization programs: the rollercoaster of AI adoption is opening up increasingly reliable options for autonomous optimization through advanced supervisory control.
This article is part of our Building Owner Signal series. We're highlighting patterns we see across conversations with building owners. These short insights focus on operational risks, emerging priorities, and what owner teams should pay attention to now.
Register for the next Nexus Labs event.
Sign up for the newsletter to get 5 stories like this per week:
This is a great piece!
I agree.