I recently purchased a virtual reality (VR) headset and used it to visit places around the world in Google Earth. I flew over petabytes of data towards South America and began searching for Huanchay, a small district 8,500 feet high in the mountains of Huaraz, Peru. The remote village has just one main road and is not an impressive showcase of VR technology. I knew there would be no street views or panoramic photos of Huanchay. I was looking for a grainy satellite image of a 600 square foot adobe house that I helped design and build 8 years ago while getting my master’s degree in solar energy engineering at UMass Lowell.
I found the house with the small interior courtyard that provided passive solar heat at the end of that one main road. When I last left Huanchay, it was little more than a foundation. Now that I’m a data analyst, how I wish I could read data from the temperature sensors inside to see how much heat is collected from that solar courtyard!
The service-learning program at UMass Lowell worked with graduate students to bring solar projects (and soccer balls!) to Peru for over 20 years. The success of the program was due in part to the continued presence in these remote areas to regularly maintain the systems and educate the communities. Without this “ongoing commissioning,” a battery could fail in a solar charging station, for example, and not only would the people be without power, they would lose faith in the technology. Routine follow-up also enabled UMass to recognize opportunities – schools that needed small solar installations to qualify for laptops from the government, asparagus farms that could benefit from wind powered drip irrigation, or a small medical clinic for pregnant mothers in Huanchay – which informed our design for the house we built. It was the farmers that UMass had been working with for years that drove us up to Huanchay in the back of their asparagus truck. It was local Peruvians who named our project, “Casa Solar Maternidad”.
It took years to design, redesign and build the passive solar house I flew over with my VR goggles. We adapted our design as our familiarity with the region grew. For example, the size and quantity of the wooden beams we used for the roof were limited by the wood we could cut from the fallen and dried eucalyptus trees in the area. We had to use corrugated plastic for the courtyard sunroof because plexiglass was not available. We were asked by the mayor to increase the thickness of the exterior walls by 6 inches to better withstand earthquakes.
The window coverings would be sewn from blankets by local women.
Some of our redesigns were scribbled on scratch paper while hiking for wood. There were many ‘change-orders’ but the result was a project that would be functional and sustainable for the doctors and mothers in Huanchay who were to use it.
Several years later, my first job out of school was developing performance contracting projects as an energy engineer. I would audit buildings and deliver a package of energy efficiency and general facility improvement measures that would pay off over time with energy savings. Fresh-faced out of college, I thought I would be evaluating hot new technologies and using advanced energy modeling programs. What I found was a lot of deferred maintenance needs – disconnected fan actuators, dirty filters, leaking steam traps, oh my! A typical performance contract was 5 to 20 years and utilized energy savings from quick payback measures like efficient lighting or controls sequences to balance higher payback measures like replacing HVAC equipment beyond its useful life. I know what the future holds for equipment without maintenance programs, and I wish I could use my VR headset to look at the work I did in those facilities to see how they are operating years later. Now that I work as a data analyst, I have a way to do just that.
When I was manually gathering and entering information over the course of weeks from equipment data loggers during audits, I had no idea how much value was hidden in the thousands of points buried in the building automation system (BAS) and how much more efficiently that information could be harnessed. As an analyst, I now have access at my fingertips not only to this seemingly overwhelming amount of data, but to powerful algorithms that turn the data into a list of operational faults and inefficient sequences of operation at the central plant, air handling, zone and utility metering level. I can monitor an entire campus of facilities with hundreds of interconnected pieces of equipment and manage faults in real-time. By communicating this information to facility personnel, work-orders are issued and faults are fixed, and I am able to track changes in the data to verify and quantify the improved performance of the equipment.
One particular project stands out in my mind as the most satisfying to date. Through data analysis, I detected an oscillating hot water heating valve. I notified the facilities manager that the valve was opening and closing rapidly. An automation technician was called in to investigate the issue and found that hot water was leaking through the closed valve at the stem. The issue was not detected at the site because the airstream temperature remained unaffected, and therefor there were no comfort complaints. Before the valve oscillation could be addressed, the valve leakage progressed to the extent that the temperature of the airstream directly after the heating coil was over 100 degrees, more than 50 degrees hotter than the intended temperature. I again notified the facility, the valve was verified to be blown open, and it was replaced within 30 days. This repair resulted in more than $173,000 in avoided annual energy costs. This project was so rewarding because my ongoing relationship with the equipment, the facilities personnel and the building automation system technician ensured the fault was caught and resolved in a short time frame.
Building equipment need constant, continuous commissioning to operate as intended, and some faults can remain undetected for months, and even years. Components fail, schedules drift, sensors lose their calibration, and equipment is put in override. Data is immensely important in identifying these faults, as are the fault detection algorithms, but just as important is an ongoing relationship with the site and personnel. Our recommendations adapt as our familiarity with the equipment, building operation, and areas of concern grows. Just as I couldn’t design a home for use in Peru from my chair in Boston by merely looking at Google Earth, the accuracy and relevancy of my building data analysis requires an ongoing relationship and understanding of the history of the site. And as my recommendations are implemented and leaking valves are fixed, I can still watch over it all without leaving my office, which is ALMOST cooler than my VR headset.
Julianne Rhoads – CEM, EIT – Senior Analyst