Skip over navigation

From Smart Grid Vision to Reality:
Agent-Based Modeling of the Smart Grid

Adviser: Alain L. Kornhauser

Pawel R. Buczak

Operations Research and Financial Engineering

“It combined every major category of skills and knowledge I had developed through my Princeton coursework: mathematical modeling, microeconomics, public policy, and programming.”


At the beginning of the summer after my junior year, I made a contract with myself to have a thesis topic and some initial research done by the time I set foot back on campus. The summer came and went, and my contract remained unfulfilled as I moved into a Princeton dorm for the last time. This failure was not for want of trying. I spent a significant amount of time thinking and searching for a theme to focus on. Being a member of the operations research and financial engineering department, I had an extremely large universe of possible topics at my disposal—statistics and optimization apply to everything, not just finance, as even some of my fellow Princetonians would have you believe. My only true binding constraint was that my topic had to have a public policy component in order to complete the requirements for my certificate in the Woodrow Wilson School of Public and International Affairs. I had many rather good ideas, but that perfect original topic that would revolutionize the world yet also allow me to enjoy senior year did not materialize even after a summer-long hunt.

The search for an extraordinary yet doable thesis topic was my first mistake and certainly not my last. Ambition is a quality that all Princeton students seem to share but, in this case, my determination to choose an exceptional topic hurt my overall thesis progress. Spend too much time deciding on a topic and you risk shortchanging yourself in terms of time for actually completing the thesis. No one understood this better than my adviser, Professor Alain Kornhauser. After a few meetings where I came in armed with possible ideas, we systematically struck down infeasible or uninteresting topics until there was only one left: the smart electric grid.

While I arrived at the smart grid through the process of elimination, in retrospect it was the logical starting point in my quest for a thesis topic. The spring before, I had written a junior paper in the Wilson School giving an overview of the benefits of the smart grid and the potential barriers to its growth as well as some policy options to induce its expansion. This was a subject I was passionate about as well as an area that could truly benefit from an original modeling analysis. I had found my topic.

In the end, my thesis dealt with the economic incentives and undercurrents that form in wholesale and retail electricity markets due to the particular characteristics of the smart grid. It combined every major category of skills and knowledge I had developed through my Princeton coursework: mathematical modeling, microeconomics, public policy, and programming. With the model I made, I was even able to explain some seemingly paradoxical results found in earlier studies as well as capture previously postulated effects not yet seen in a model of real time electricity pricing.

And yet, while I consider my thesis to have been a tremendous success, this is not to say that there were not many potholes along the way. How you deal with these roadblocks defines the eventual success or failure of your thesis. I cannot overemphasize the importance of continually chipping away at the rock that is the thesis even though it may not show any signs of cracking. Unable to find some particular data set? Temporarily move on to writing the code for your model. Stuck debugging? Start writing your literature review. Got writer’s block? Compile your bibliography or table of contents. There is always something to accomplish, and time is too valuable to be spent getting frustrated instead of moving forward.

Don’t get me wrong—I am not implying that you should be living in Firestone Library for the duration of your senior year. Far from it. Time management is crucial, and that means making some kind of progress when it is designated thesis time while also cordoning off time for other classes and fun with friends. In my case, as I was on the men’s varsity basketball team, this also meant allotting many hours for time-consuming practices, workouts, and games. Setting particular days and times to work on the thesis—and constantly compelling yourself to accomplish something in these hours—is a good way to stay on top of the process. Concrete deadlines throughout the year like those established by Professor Kornhauser also helped the work from snowballing to an unmanageable size.

I believe that switching the sections of the thesis I focused on helped me successfully manage my time effectively as well as minimize my frustration. However, I found that this strategy also improved the quality of my thesis in other ways. Writing about my model as it still was not operational forced me to revisit minute details that spawned little epiphanies on how to expand the model and improve its design. The model I ended up with in April was markedly different from the early outlines sketched out in September. In general, writing during the course of the research and planning stages is always relatively more effective because the material’s specifics are fresh in your mind.

While all these tidbits are hopefully helpful, the most important aspect of the senior thesis is that it is a highly personalized, constantly evolving process. Every thesis writer comes away with different experiences and insights, most of which are completely unexpected. This may seem daunting at first, but do not attempt to impede the evolutionary nature of the thesis process; rather, harness this energy to fulfill your greatest intellectual journey.

From Smart Grid Vision to Reality:
Agent-Based Modeling of the Smart Grid

Pawel R. Buczak

Alain L. Kornhauser

Professor of Operations Research and Financial Engineering

“If it’s academically rewarding and intellectually challenging, and they are passionate about it, we should pursue it, and each weekly meeting will be lots of fun.”

In each of my more than 30 years at Princeton, I have had the pleasure of supervising at least a few senior theses. I find it a most academically and intellectually rewarding experience because of two simple yet fundamentally important characteristics of the senior thesis process. First, the choice of topic and focus is squarely on the senior. I have dozens of pet projects that would benefit dearly from fresh and thoughtful consideration by talented seniors, but they’re what I love. The thesis is an opportunity for seniors to pursue in depth what they love. Their internal fire and enthusiasm for the topic essentially ensures a successful process and outcome.

Second, there is no need for the project to pass muster with some outside funding agency, employer, donor, or corporate entity; thus, this is an opportunity to investigate even “wild and crazy” things. I challenge students to consider elements and aspects of their topic that are not necessarily in the mainstream—elements that others would consider unfundable. If it’s academically rewarding and intellectually challenging, and they are passionate about it, we should pursue it, and each weekly meeting will be lots of fun.

While my field is transportation and Pawel had taken my “Transportation Systems Analysis” course, Pawel’s passion wasn’t transportation, but energy. Energy has grids, transportation has networks, so I wasn’t so far away. Also, the inception of Plug-in Hybrid Electric Vehicles (PHEVs) could have substantial implications for the future intelligent generation, distribution, and pricing of energy. In fact, we were very closely aligned. In the process, he made very important findings by projecting the anticipated consumer responses to the dynamic pricing of electricity as influenced by various adoption rates of PHEVs, their use as a distributed storage medium, and consumer investment in household-sized renewable (solar and wind) energy-generation devices. Pawel also had his “wild and crazy” subchapter investigating the technical feasibility and safety of what he termed “neighborhood nukes”: small-scale (serving about 10,000 homes) distributed nuclear fission power plants. In the end, Pawel’s thesis is a substantial intellectual contribution with substantive public policy implications to the regulation of public utilities and the promotion of clean energy generation.

Moreover, it lives on. Inspired by Pawel’s thesis, Stephanie Lubiak ’11 is pursuing in depth the critical technological, environmental, safety, national security, geo-political, and private-sector economic aspects of neighborhood nukes as her senior thesis.

While passion and enthusiasm are important, the thesis also takes time. In some sense, the sooner one can start the better. In operations research and financial engineering, we have instituted a process that begins right after midterms of the spring semester of the junior year. The faculty describes to the rising senior class the broad topic areas that encompass their expertise and interest. Through a series of meetings, students prepare an ordered list of topics and advisers. Matches rarely evolve to lower than a second choice. To start early, students are encouraged to perform background research and preliminary data collection during the summer. Then, two weeks into the fall semester, each student is required to submit a title, preliminary abstract, proposed table of contents, problem statement, background literature survey, and initial bibliography. While each of these can and will most likely be changed in the final document, this milestone ensures that each student hits the ground running at the beginning of the fall semester. Several interim versions maintain progress throughout both semesters. A traditional champagne toast and wading in the Wilson School fountain follows immediately after the final deadline. In the end, returning alumni recall it as their most memorable and rewarding academic experience. That’s why I love playing my part.