In India, we are now becoming accustomed to air-conditioning and refrigeration units for our day-to-day activities. Once a luxury, air-conditioners (and refrigerators) are much more affordable; thus, steadily turning into a necessity, especially during the summers. Accordingly, 70% of future demand for air-conditioning and refrigeration units is expected to come from 2.8 billion people living in the hottest emerging economies of the world (including India), where currently only 8%-10% of people own cooling and refrigeration units in their home. With that being said, are we ready to deal with such a massive demand? The United States of America, where 90% of people are already using air-conditioning and refrigeration unit in their houses, is wasting some 212-billion kWh of electrical energy every year . As per a report by American Indoor Greening Institute, “On a per home basis, the wasted energy (by Heating, Ventilation, and Air-Conditioning) is at least five-times, and in some cases, up to twenty-times more than the combined use of all other home devices.”.
Irregular maintenance, improper operations, and negligence are some of the many reasons that result in frequent breakdowns of an appliance. While some failures are sudden, others happen gradually over time, also known as slow time-varying faults (or degrading faults). Often, we will spontaneously identify a sudden breakdown and call a technician to repair the appliance; but, we usually fail to sense the early symptoms of slow time-varying faults. Refrigerant leakage is one such slow time-varying mechanical fault, quite common in air-conditioners and refrigerators. As the name suggests, refrigerant leaks through a puncture hole that often starts as a pinhole leak and becomes bigger when goes undetected. Due to the loss of coolant, the appliance works with reduced efficiency and takes more time than the usual to cool the space, thus, wasting significant energy. In addition to energy wastage, the leakage exposes tenants to refrigerant which is extremely dangerous for their health.
The consequences of refrigerant leakage are even worse for retail outlets who set up cold-rooms to preserve perishable food items (usually at 5 – 8 degrees), and the stored product goes bad due to improper cooling by the refrigeration unit during the leakage. In Figure 1, we depict one such instance of refrigerant leakage from a retail store in which refrigeration unit (akin to an air-conditioning unit) broke one week after the leakage started. Just one day before the complete shutdown (on March 30), though the refrigeration unit was consuming significant energy, it was hardly cooling the space with room temperature being more than 25 degrees. In addition to repairing cost, the owner also dealt with the staling of stored products and business loss owing to the downtime.
Now, you might be wondering, what is the solution? Definitely, you don’t want to keep the appliance running when it is leaking, because that would – (1) waste significant energy, (2) expose you to hazardous refrigerant molecules, (3) keep the room temperature out of desired range, (4) and damage the appliance. One way is to use leakage detectors (such as Halide Leak Detector) regularly; however, there are two main concerns with this approach. First, one must use them with certain precautions. For instance, we cannot use Hailey detector with hydrocarbon refrigerants. Second, even if you are well trained to handle such leakage indicators, you might need to use these detectors on regular time intervals; and that is cumbersome. I suppose using a leakage detector at the end of each day is not a very user-friendly way to handle the problem. Therefore, we designed a framework – Greina – that only uses room temperature and occupancy information from a smart thermostat to timely notify you about such leaks.
Greina is an Icelandic word that means to identify. Greina takes ambient information from a smart thermostat and weather conditions from a cloud-based weather server to tune the parameters of a lumped thermal model. The tuned thermal model simulates the room temperature, and Greina then compares simulated room temperature with the sensor measurements. When the actual temperature is sufficiently above the simulated room temperature, the system raises a red flag. Of course, not every red flag is a leakage flag. Room temperature can be higher than the estimates even when the refrigeration unit is working fine (false positive), and vice-a-versa, the framework might confuse the initial symptoms of leakage with the noise generated from manual interventions (false negative). For instance, in Figure 1, the room temperature is in the range of 8 – 10 degrees in the initial stages of the refrigerant leak, and that can also happen because of occupants’ activities within the space (as shown in Figure 3).
What could be the consequences of such a misclassification? Well, that depends! If misclassification is a false negative (misinformed that the refrigeration unit is fine), then repairing will get delayed until the outlet manager identifies the leakage. However, if misclassification is a false positive (misinformed that the refrigeration unit broke), then the company might end up paying a significant amount to their maintenance contractor for the unnecessary visits. While such visits are immoderate, they are annoying for the store manager and disruptive to the daily operations. Henceforth, Greina employs CUSUM (Cumulative Sum Control) technique to ensure that the user gets notified only when the system is confident about the leakage.
Basically, for each room r, Greina maintains a bucket, (b_r), to gain confidence before reporting leakage to the users (as shown in Figure 4). Whenever room temperature goes beyond the decision boundary, Greina increments the bucket value by one. For every consecutive hour, when room temperature is within the expected range, the bucket value decreases by one unit. Once the leakage is detected, if the room temperature stays below the decision boundary for some consecutive hours, Greina assumes the refrigeration unit is working fine and resets the bucket. In case of missing information, the bucket value remains unchanged. We evaluated Greina using the data collected from 74 outlets of a retail enterprise for one year. Our results indicate that Greina is comparable or better than the current techniques of leakage detection. It can reduce average reporting delay by a week with best of around 20-30 days in few instances.
Our analysis indicates, that during these days, the retail enterprise could have saved twice the energy a refrigeration unit consumes on a typical day. Moreover, by timely repairing the refrigeration unit, they could have kept a room 5 – 10 degrees colder every day, when the refrigerant was leaking. Thus, in addition to being scalable, our study indicates that Greina is reliable and effective in real-world. Since Greina requires minimal (or no) intervention from the store manager, it also ensures non-interruptive working hours. We have documented many such interesting insights (about Greina and the study) in our paper titled, “Beyond Control: Enabling Smart Thermostats for Leakage Detection”, published in Volume 3 Issue 1 of the Proceedings of the ACM on Interactive, Mobile, Wearable, and Ubiquitous Technologies. You may find a preprint of the paper in the publication section of my webpage, or access it through the arXiv. Please feel free to share your feedback, suggestions, comments, or queries in the comment section below. You can also contact me directly through E-mail. The contact details are available on my webpage.