Start With Better Data: What UW Learned From Its Legacy Chilled Water System
Before replacing legacy chilled water infrastructure, the University of Washington first asked what its existing system could tell it.
By Scott Usselman, Nate Mostajo, and Joe Wilkinson
Across higher education, campus facilities leaders are being asked to do more with infrastructure that was built for a different era. Central utility plants and chilled water systems must support growing demand, sustainability expectations, and tighter operational needs, often while major capital replacement dollars remain limited or years away.
At the University of Washington, that familiar challenge came into focus within an aging chilled water system, served by two central plants (one modern and optimized, one legacy) serving a large and dynamic campus. The system itself had not changed dramatically, but the demands placed on it had. Facilities teams needed better information to operate the legacy plant efficiently while meeting increasing cooling loads and maintaining reliability during the cooling season.
Rather than beginning with an assumption that the answer was wholesale equipment replacement, the project team focused first on a more fundamental question: What can we learn from the system we already have? The resulting work offers a practical lesson for other campuses: before replacing legacy infrastructure, it may be possible to unlock meaningful value by improving measurement, visibility, calibration, and operator confidence.
Understanding the Campus Utility Challenge
Like many institutional central plants, UW's chilled water system is operating within the realities of age, complexity, and limited visibility. One of the plants serves the campus loads through multiple distribution tunnels, yet the team did not have a complete picture of how much chilled water energy was being delivered through each pathway from this plant. Operators knew how much water was leaving the central utility plant, but not how many BTUs were moving through each of the five tunnels serving campus.
That lack of visibility made optimization difficult. Without reliable tunnel-level flow and temperature data, the team could not confidently evaluate where energy was going, how loads were shifting, or which operational adjustments might have the greatest impact. The challenge was not simply that the infrastructure was old; it was that the system did not yet provide the level of information needed to manage it as strategically as campus needs required.
Looking Beyond Replacement
One obvious path would have been to focus immediately on chiller replacement. For many campuses, that is often where conversations about aging chilled water infrastructure begin. New equipment can be necessary and appropriate, but it is not always the first or only answer. In this case, the project team believed there were lower-cost actions that could produce significant operational value before major capital decisions were made.
The near-term opportunity centered on efficiency and reliability. If the campus could improve the accuracy of its sensors, establish trustworthy energy measurements, and give operators better information, it would be better positioned to operate the existing plant more effectively and plan future phases from a stronger factual foundation. For APPA member institutions facing similar capital pressures, that distinction matters. Optimization does not eliminate the need for renewal planning, but it can create a practical bridge between today's operating realities and tomorrow's capital investments.
Improving Visibility and Trust in the Data
The first phase focused on instrumentation. The team replaced 26 temperature sensors, one differential pressure sensor, and seven flow meters, while adding four additional flow sensors for the campus tunnels. Those improvements changed the quality of the data available to the facilities team. Sensor accuracy improved from errors of up to 10 percent to less than 3 percent, giving the team a much stronger basis for understanding plant performance.
Just as important, the new tunnel flow sensors made it possible to understand how much energy was moving through each tunnel. In chilled water systems, BTU measurement depends on accurate flow and temperature data. If either input is unreliable, the resulting energy picture can be misleading. By improving both the quality and granularity of the measurements, UW created the foundation for better operational analysis and future control strategies.
Operator involvement was essential. The operators helped check calibration of the new sensors and identify issues that emerged along the way. The team also built confidence by showing operators, in real time, how the new readings compared with calibrated instruments. That process matters because automation cannot succeed on data alone. Operators need to trust what the system is telling them before they can rely on it to support decisions.
Preparing for Smarter Controls and Operations
Although the long-term opportunity includes controls, automation, and system-level strategy, the project is still in the discovery phase. No major plant staging or chilled water distribution system balancing changes have been implemented yet. That is an important part of the story. The team did not leap directly into automation. Instead, it began by making sure the underlying information was accurate enough to support future decisions.
For campus facilities leaders, this sequencing is instructive. Reliable automation depends on reliable inputs. A sophisticated control sequence built on questionable data can create confusion, erode operator trust, and make performance harder to diagnose. Because UW's district energy system is served by both a modern central plant and a legacy plant, the team is no stranger to highly automated and optimized operations. The opportunity is not simply to introduce automation, but to extend greater visibility, consistency and control across a complex system with different generations of infrastructure. By starting with instrumentation, calibration, and visibility in the legacy plant, UW positioned itself to evaluate future operational strategies with greater confidence over the next six months and beyond.
The work also had to be completed around an operating chiller plant in the middle of cooling season. That constraint required coordination with plant staff and reinforced the value of practical, phased improvements. In many campus environments, the most useful solutions are not the most disruptive ones; they are the ones that can be implemented while mission-critical systems remain in service.
Practical Lessons for Other Campuses
The immediate result is a chilled water system that can be measured with far greater confidence. Before the sensor replacements, some readings included substantial noise and variability. After replacement, the new sensors produced cleaner, more stable data. The difference may sound technical, but its operational value is straightforward: better data helps facilities teams see what is actually happening, rather than what they assume is happening.
That visibility is especially valuable as campuses pursue energy, carbon, and resilience goals. If a facilities team can see where chilled water energy is going, it can ask better questions: Which tunnels or zones are driving load? Where might operational changes improve efficiency? Are future capital investments targeting the right problem? What should be fixed now, and what can be planned for later?
The project also offers a useful reminder about the role of operators in technology-enabled facilities work. Instrumentation and automation are not separate from people; they depend on people. UW's operators helped validate calibration, troubleshoot field conditions, and make the new data more usable. Their participation helped turn a technical upgrade into an operational tool.
For institutions evaluating legacy chilled water systems, the takeaway is not that replacement is unnecessary. Some assets will need to be renewed. The lesson is that replacement should be informed by accurate system knowledge. Before buying new chillers because equipment is old or a sales recommendation suggests it, campuses can benefit from asking what can be improved with the infrastructure already in place.
For campus facilities leaders, that perspective is both practical and strategic. Better sensors will not solve every infrastructure challenge. But better measurement can reveal which challenges matter most, help teams prioritize limited resources, and create the confidence needed for smarter operations. In an era when campuses must do more with existing assets, that may be one of the most valuable first steps.
About the Authors
Scott Usselman is a principal, and Nate Mostajo and Joe Wilkinson are senior project managers, with EEI Building Performance.
Connect on LinkedIn: Scott Usselman and Joe Wilkinson
Read the Full Issue
This article first appeared in the Fall 2026 issue of Facilities Manager, the magazine of APPA. Read the full Fall 2026 issue for more stories from member campuses.
The Magazine of APPA
Facilities Manager, Fall 2026
APPAU St. Louis, Start With Better Data, AI-Ready Workforce, and The Most Important Call.
Have a project or lesson to share with your peers? Email publications@appa.org to submit to Facilities Manager.