The Chiller Plant as Living Lab: What High-Tech HVAC Systems Can Teach Students

This article originally appeared on The NEWS
By Bert Valdman, CEO of NorthStar Energy

Facilities leaders faced with curtailing rising costs and meeting sustainability targets while keeping building occupants comfortable and maintaining climate control for labs and other sensitive spaces need increasingly sophisticated systems to achieve their goals. These systems justify themselves based on hard-dollar returns on investment. But looking at them only through that lens misses a significant opportunity to contribute to the institution’s educational mission.

Campus physical plants that have become internets of things — requiring cloud-based management, machine learning, and visualization tools — could serve as more than unseen controllers of the classroom environment. They could be the classroom. Mechanical engineering students could explore the plant with managers and engineers to see how a cutting-edge HVAC system works, for example, and visualization tools could bring the plant into the classroom. The trove of data these systems produce could also give data science students meaningful analytics projects.


What will students see when they look inside an advanced campus HVAC system? What could they learn about?

At Georgia Tech, which hosts class visits to its chiller plants, students can get a look at two large, optimized chiller plants with mixed-equipment systems—a common situation. One plant has seven chillers, seven condenser water pumps, seven cooling towers, and one free-cooling heat exchanger, and the other has seven chillers, seven condenser water pumps, and three cooling towers.

Technology includes a state-of-the-art chiller that can handle wide variations in water flow, a water- and energy-saving pump configuration, and an advanced optimization software implementation tailored to meet rigorous network security requirements while providing the continuous support and monitoring needed to maintain energy and cost savings. Georgia Tech and Optimum Energy worked together to develop a way for Optimum Energy’s cloud-based OptiCx solution to function without direct network access. Every five minutes, data passes from the BAS to data loggers, which OptiCx reads through a VPN-protected database. Now the Georgia Tech team can see and evaluate system performance in real time.

The University of Texas at Austin demonstrates a multiple-building, data-driven HVAC implementation. The campus optimized four chilled water plants (totaling 45,000 tons) on a common loop with a 4 million gallon thermal energy storage (TES) tank in a series of projects that followed the three laws of optimization:

  1. What cannot be measured cannot be optimized;
  2. Optimize systems, not just individual components; and
  3. Optimization must be automatic, dynamic, and continuous for maximum efficiency.

The project included control of chiller staging for all four plants, variable differential pressure control and flow control, and TES charge and discharge. The university is now saving 21,000,000 kWh, 200,000 MMBtu of steam usage, and 4 million gallons of water annually.

Learning opportunities depend on the plant setup, but the chance to experience a plant before and after optimization, or to compare optimized and unoptimized plants, would be enormously valuable. The monitoring data and performance statistics an optimization platform produces provide plenty of ways to explore the factors in plant efficiency.
Even newer plants have teaching potential. We worked on a university bioscience research facility plant that was only five years old and had inflexible climate requirements. Post-optimization, the plant runs 27 to 37 percent more efficiently and is a living demonstration of maximizing efficiency. The university’s director of facilities and lab services converted it to an all-variable flow plant and then added an optimization and control layer. From the variable speed drives and sensors installed on chillers, pumps, valves, and tower fans, the software collects a tremendous amount of data about the plant equipment, including water flow, electrical power consumption, load conditions, and more. It compares the data to control algorithms, assesses plant conditions in real time, and then automatically changes pump and fan speeds using chilled water temperature, equipment staging, and other operational changes to maximize efficiency.


Last year I taught an entrepreneurship class with two other clean-tech executives at the Institute for Sustainability and Engineering at Northwestern University, and it was clear that students crave hands-on experiences with the latest technology and need to understand real-world applications.

Most colleges and universities are not yet bringing students into their chiller plants and boiler rooms when they upgrade, but they’re missing a real opportunity. One of our engineers, a former campus facilities manager, brought in mechanical engineering students as interns and found the experience mutually beneficial. “They bring a lot of enthusiasm, and it helps them to put theory and practice together,” he says, adding, “It’s the greatest way to hire new staff. The benefit for the university is obvious.”

Colleges may even benefit from new ideas from students—our engineer also sponsored an annual senior project at the engineering school, and ended up implementing a couple of ideas the students came up with. We’ve seen the benefits of working with students ourselves. At Optimum Energy headquarters in Seattle, we worked with University of Washington students on a machine learning project, and found that both sides benefited. As one of the students told a local radio reporter about the program, “Oftentimes when you take a class everything’s super neat, everything works out super nicely. But that’s not the real stuff. In the real stuff, there’s actual work to be done,” like discovering a faulty sensor that messes up your datasets.

Based on our experiences, we will gladly have engineers work with students on-site at our installations. We encourage facilities directors at colleges and universities to invite students to learn in the chiller plant and boiler room: they’ll be contributing to their institution’s educational mission and gain a new perspective on their plant.

Schools that view their HVAC plants as part of the educational experience can engage their students with the next generation of building technology in a way that gives them the insights needed to develop ever simpler, more powerful, and more cost-effective building systems. They’ll know what a high-performing plants looks like, because they’ll have seen one in action. That’s an ROI that will benefit the entire world.

Published November 19, 2018

Want More Smart Buildings? Amp Up the Collaboration

This article originally appeared on

By Bert Valdman, CEO of NorthStar Energy

“Smart technologies are defined by their interconnectedness,” points out a recent ACEEE report on smart building markets. The companies that buy and sell them, however, are defined by their disconnectedness.

Intelligent buildings have been a concept for decades (a quick search turns up a research report from 1991). Sustainability thinkers have been advocating them for years, and they’re a hot topic in building trade publications. So why is it that the brightest thing about most buildings remains their always-on lights?

There are many answers to that question: misaligned incentives, lack of accountability for energy costs, the fact that some professionals in the building maintenance sector see standardization and automation as threats to their livelihoods. Those are all topics of discussion, if not sufficient action. But there’s another core issue that’s been largely overlooked: every actor in the smart buildings universe is an island. Here are some of the culprits:

  • Corporations looking to control their energy destiny and improve sustainability often aren’t structured to drive action across the enterprise. Energy-related decision making is decentralized, and local purchasers typically lack the time and expertise needed to make sense of an onslaught of new technologies.
  • Emerging companies in smart building and energy technologies burn through precious capital waiting out the resultant long purchasing cycles and building large sales and business development teams to reach the many levels of decision makers.
  • These teams all target the same commercial and industrial customers with siloed solutions.
  • Established energy service and building technology conglomerates know customers want integrated, comprehensive solutions, but they’re structured to market individual portfolio company products and often hesitate to cross organizational boundaries.

One way to break down these boundaries is to adopt a collaborative business model akin to those sustainability leaders have used to advance fair trade and resource conservation.

What Does Collaboration Look Like?

A shared services organization that is customer focused and solutions oriented, and that receives active support from commercial customers seeking integrated solutions and willing to provide needed data, could be the answer. This model would enable emerging companies to go to market more efficiently and could incorporate a vetting component that makes technology capabilities and comparisons transparent for corporate buyers.
We can take a cue from the food world, where various forms of pre-competitive collaboration are increasingly common. Industry organizations in coffee, chocolate, seafood and other sectors bring together key players in the supply chain—importers, processors and retailers, say—to support projects that ensure a sustainable, high-quality supply chain that benefits all participants. (See “It’s all hands on deck to save seafood supply chains” for a few examples.) Smart building owners and vendors can achieve similar results by working together as an ecosystem.

Many of the products and solutions that smart building technology companies offer are complementary. These companies spend a considerable amount of time and money on marketing and sales, going after the same customers at the same companies. A highly skilled and trusted shared services organization that marketed members’ products in a fair way would get technologies and services to market more efficiently.

It would also address barriers that prevent large corporations from moving forward on energy initiatives: complexity, a lack of familiarity with the technology, and trust. A shared services organization could develop a checklist of everything in a building that uses energy or water; technologies and practices that will make building systems maximally efficient; and specification, purchasing and use guidelines.

In concert with this planning and education tool, the organization could present integrated solutions that meet an enterprise customer’s particular financial, operational, and sustainability goals—for example, a suite of lighting, HVAC optimization, and demand management technologies that shrink the electric load; distributed renewable energy resources that reduce each facility’s dependence on the electric grid along with its carbon footprint; and a control dashboard that orchestrates the whole thing while providing business intelligence.

Who should lead this effort, or something like it? The utility industry is ideally positioned to take this on (and I say this as a former utility executive): they run the backbone energy systems and they have relationships with technology providers and commercial users. Enterprises could also lead by actively seeking integrated solutions, encouraging collaboration among technology companies and service providers, and materially supporting a collaborative approach.

Think Forward, not Backward

It’s easy to come up with reasons a collaborative approach wouldn’t work, but they’re all based on a status quo mindset positing that something can’t happen because it hasn’t already. The model could take off if we start with an initial commitment to not waste time creating elaborate contractual structures focused on “what if it doesn’t work?” We can find ways to maintain each participant’s intellectual property. If we don’t bog it down from the beginning with nuclear disarmament–level negotiations, we can create a collaborative structure that benefits emerging technology companies, energy conglomerates, corporate enterprises, and the world at large.

A collaborative effort can scale faster, deploy technology faster, and drive innovation into the DNA of an organization faster. If we are willing to work together, we can create the intelligent buildings we’ve all been seeking, but somehow always remain in the future. We just have to care enough to invest the effort.


How to Make Buildings Truly Smart – and Keep Them from Losing Their Minds

This article originally appeared on

By Bert Valdman, CEO of NorthStar Energy

We all hear a lot about intelligent buildings – so much that you’d think they were everywhere. But when you look at the data, it seems that most buildings are as dull-witted as ever.

Energy efficiency is a core aspect of a smart building, and yet building energy use has risen over the past several years in even the most efficiency-conscious cities, according to an analysis of data from the American Council for an Energy-Efficient Economy. Technology is not the problem. U.S. commercial buildings could cut energy use 29 percent on average by taking full advantage of controls technology and implementing a few other basic energy efficiency measures, a study by the Pacific Northwest National Laboratory found.
The United Nations, noting that about 40 percent of today’s global greenhouse gas emissions come from buildings, is taking on the problem with its new Global Building Network. The network, with Penn State’s Institutes of Energy and the Environment as a lead institution, aims to create an international framework that will make buildings more sustainable, more efficient, and healthier to live and work in.

Standards are clearly important. Issues like ineffective controls and a mismatch between design assumptions and building occupant behaviors deserve plenty of scrutiny. Part of the problem, though, is the persistent issue of performance drift – buildings and building systems should be highly efficient, but performance deteriorates rapidly or never matches the model. To really solve that problem, we need to rethink the metrics we use to measure energy design for smart buildings.

Accountability Drives Efficiency

Singapore provides an excellent model. It has been rolling out a policy that builds in accountability for meeting carefully crafted performance targets, and the city is already seeing significant success. Singapore started with a standard for cooling systems based on metrics for efficiency rather than gross energy use. This puts the responsibility on the designers: a cooling system can meet an efficiency-based performance standard even if building operators overcool. If the building doesn’t meet that standard, you know the problem is the system, and if it does but still uses too much energy, you know the problem is operational.

As for drift, advanced high-performance systems require monitoring and support to function optimally over time. Singapore addresses this issue by requiring that the system’s design and construction include measurement and verification features. Initial results from Singapore’s shift in metrics are dramatic: cooling systems in existing buildings subject to the standard are demonstrating an average efficiency improvement of nearly 50 percent.

Now Singapore is moving on to overall energy use indices (EUIs) for various building types, along with a simple way to track energy performance. The idea is that with all these guidelines in place, it will be easy to see where performance problems are coming from – the system or its operation – and the route to a more efficient building will be clear.

More Ways to Inspire Action

Another way to encourage accountability is to establish a metric for building efficiency that is relevant for all types of facilities – weather-adjusted BTUs per square foot, for example. That would allow easy benchmarking: publicly rating facilities within their category (data centers, hospitals, office buildings, and so on) would give designers and operators a target and a motivation for working to make them top performers.

Linking energy modeling to energy action would also lead to better performance. Architects now have tools that let them model energy usage in any number of scenarios to see where the greatest savings are. Similarly, building operators can visualize their energy use and the effects of various actions through dashboard tools, but they often don’t take action because doing so seems too complex or disruptive. These visualizations need to be tied to automated optimization software that takes the friction out of energy efficiency decisions.

Visualization and action are symbiotic. Unless dashboards drive actions, they will always be one-dimensional. Conversely, if all you’re doing is acting, you won’t be able to accurately assess the value of your actions or see when things go off track.

Designing for Today’s Needs, Not Yesterday’s

Real building efficiency also requires reimagining building infrastructure with focus on sustainable use. Building components that use the most energy, such as lighting and HVAC systems, often are designed, configured, and used based on the habits of decades past. When that happens, even systems that incorporate the latest technology will use more energy than necessary.

The biggest challenge with HVAC systems – which account for nearly 50 percent of the typical commercial building’s energy use – is that they are more highly engineered and complex than they need to be. They’re also sized to meet maximum peak energy needs, not for efficiency. In the retrofit market, the best we can do is improve poorly designed systems. That’s well worth doing in terms of efficiency results, as the Singapore example shows.

But when designing new buildings, we can and should be asking: What would the perfect HVAC system look like? What would the perfect lighting system look like? How can all the building’s systems connect holistically to eliminate waste? That’s how we’ll get a technical answer to the performance drift problem, along with truly smart buildings.


avgle kita is a cum slut.