Transformers currently operating in the electric grid now have an average age about 15 years beyond designed lifetimes. Grid transformations that rapidly increase transmission capacity are needed over the next few decades to permit the increased levels of renewable energy sources necessary to support decarbonization at scale. This will require additional transformers, but supply chains are already unable to keep up with current needs for replacements and new installations. Thus, it is essential to maximize the lifetime and value of deployed transformers as much as is safely feasible. Grid-edge sensor systems that can accurately monitor transformers for operational conditions, analyze sensor data to determine transformer health, detect targeted events rapidly enough to enable initiation of autonomous control actions to prevent failures, and communicate relevant information in near real-time would help grid operators keep deployed transformers operational as long as possible. Grid-edge intelligence could identify problems as they develop, letting grid operators prioritize preventive maintenance of specific transformers. But with over 50 million distribution transformers in the U.S. electric grid, such systems must be very low cost for ubiquitous use to realize the desired benefits. Existing industrial power monitoring sensor systems are prohibitively costly for this use, given the capital cost of the equipment being instrumented. The most important parameters for determining distribution transformer health are transformer loading and transformer temperature1. Overloading and unbalanced loading can lead to winding currents that increase the internal temperature, resulting in accelerated aging and reduced service life. The proposed effort will develop a low-cost sensor system to monitor current, voltage, and temperature for in-service distribution transformer health monitoring. This system will incorporate grid-edge analytics and rapid data backhaul. The proposing team has adapted cell phone filter technology to produce lightweight power line sensors that measure current, voltage, and line temperature wirelessly and without batteries on 25kV-class lines, and low-cost clip-on sensors to monitor power consumption in industrial equipment. The approaches used, with modifications, are applicable to the proposed effort. We propose lightweight, low- cost, easy to install current and non-contact voltage sensors with current measurement capability that clip onto transformer conductors, wired to a small data processing and communications module that mounts on the transformer exterior. Prototype system testing in Phase I will demonstrate technical feasibility and position the system for subsequent rapid commercial productization. Monitoring of distribution transformers with low-cost grid-edge sensor systems will enable utilities to extend the useful life of existing transformers and optimize maintenance activities to address those most likely to fail. This will help to maintain grid stability with increasing levels of renewable sources, electric vehicles, and energy storage.
