The solar microgrid our student chapter installed on a village school worked beautifully until the day three houses connected their own panels, and I learned more from the resulting instability than from two years of coursework. Nobody had modeled what happens when a passive grid suddenly has distributed generation pushing back — voltage rose past limits, our cheap inverters tripped, and the school lost power on a sunny afternoon. Diagnosing that taught me that the hard problem in electrical engineering is no longer generating power but integrating it, and it aimed me at graduate study in exactly that.
My foundation is in power systems — circuit theory, power electronics, and control, sharpened on that microgrid rather than only in the lab. My final-year project modeled voltage regulation for a small distribution feeder with rooftop solar, and I validated my simulation against the actual instability data we had logged from the school, which is the most satisfying thing I have done in engineering: a model that explained a failure I had watched happen.
I am applying specifically toward grid integration of renewables — the distribution-level power-electronics and control problems that our microgrid crashed into and that India's rooftop-solar boom is about to hit at national scale. This is where the field's hardest current questions are, and where my accidental education already began.
My goal after the degree is to work on distribution-grid modernization — the inverters, controls, and standards that let a grid absorb distributed generation without dropping a village school's power on a sunny day. A tripped inverter started this. I want a career making sure the next one holds.
For three years I have designed embedded control firmware for motor drives, and I have hit the wall between the code I write and the power electronics it commands. When one of our drives kept failing thermally in the field, I could instrument and log the fault perfectly and could not tell you whether the root cause was my control loop or the converter topology underneath it — because that topology is exactly the layer my job stops at and my formal training never reached. A master's in power electronics is my move to cross that wall.
My embedded work is a strong foundation. I own the motor-control firmware for our flagship drive, I implemented the field-oriented-control loop that cut our audible noise complaints in half, and I built the fault-logging system that made the thermal failure diagnosable at all. But every hardware question — converter design, thermal management, magnetics — I currently escalate to colleagues, and I am tired of being fluent in only half of my own product.
I want the graduate coursework in power-electronics design, converter topologies, and thermal modeling that would let me own the full stack from control law to switching device. The thermal failure I could log but not explain is precisely the kind of problem this training addresses — the interface between control and hardware where my current knowledge has a clean, frustrating boundary.
My goal is to become a power-electronics design engineer working on EV traction inverters or industrial drives, fluent from firmware to switching device rather than expert in one and dependent for the other. I have spent three years commanding power hardware I couldn't design. I am applying to finally understand what my own code is switching.