Posted: October 30, 2024
While it’s easy to recognize how infrastructures from roadways to cell towers support our lives, it’s just as easy to overlook how interconnected they are. Take water and energy, for example: while each system has its own requirements and benefits, they also work together to deliver services to our homes and offices—water pumps need power to work, for example. That interdependence offers engineers some opportunities—and some challenges. One engineer looking at these interconnected systems is Saskia Putri, a third-year graduate student working on interconnected infrastructure issues with Farrah Moazeni, Assistant Professor of Civil and Environmental Engineering.
That interconnected perspective makes Moazeni’s CONCISE (InterCONnected Critical Infrastructure Systems Engineering) lab the perfect fit for Putri. The lab researches water and energy systems with a focus on interconnected critical infrastructure systems and how they affect communities. Examples include the treatment of runoff water and wastewater, desalination design, biofuel and renewable energy production, and other issues.
In a sense, interconnections—the human kind—brought Putri to Lehigh. After earning an undergraduate degree in environmental engineering, she continued on to a master’s degree in the same area at Penn State, specializing in water systems, mainly in wastewater. She worked with Moazeni, who then taught at Penn State, during the master’s degree, and the two kept in touch. When Moazeni, who came to Lehigh in 2021, had an opening focusing on interdependence of critical structures in her lab, she reached out to Putri. Putri says, “I quickly applied!” she says. “Because my background is in water systems, I start with water then combine it with power systems. Water distribution systems are heavily linked to energy consumption because they rely on constant supply of electricity to distribute water, particularly for running pumps continuously.” This multidisciplinary characterizes Putri’s work, Moazeni says. Such approaches that pull on engineering of water, electricity, and computing analyze the workings of real systems and use multiple approaches to solve problems that affect peoples’ lives.
Moazeni’s lab appealed to Putri for another reason: she prefers hands-on research. In addition to reading current studies and considering gaps in the research, she says, “I also want to do something you can see, you can build, you can create.” For example, during her master’s program, she focused on wastewater treatment and built a lab-scale wastewater treatment system. And today, she is working with a quadruple tank system provided by the CONCISE lab, set up next to her laptop in the STEPS lab. About the size of a conference poster, the system features four transparent water tanks interconnected by tubes, pumps, and valves, with a reservoir underneath, resembling a simplified water distribution systems. This system is renowned in research communities due to its complex, nonlinear, and multivariable dynamics. Putri says it’s designed so that the water levels in each tank are interdependent via the interconnected tubes and pumps, forming a complex and cross-coupled system where adjusting one tank affects the others. Even at its small laboratory scale, this setup allows her to test the robustness and effectiveness of the control strategies and mimic the interdependent nature of real-world systems.
Putri divides her time between her own research for her dissertation and working on research connected with a Department of Defense (DoD)/Navy project with Javad Khazaei, Assistant Professor of Electrical and Computer Engineering. Half her week (Monday to Wednesday) focused on her dissertation proposal, which she plans to defend in February 2025. After it’s approved by her committee (Moazeni, Khazaei, Ethan Yang, Paolo Bocchini, Pietro Tesi of the University of Florence), she’ll start writing. In her research, she’s working with model predictive control, focusing on “data-driven control for critical infrastructures, specifically for water energy.” The data-driven aspect uses only available data to identify the dynamic models of the system on the power or water side. Putri focuses on the microgrid, either islanded or grid-connected with microgrids, and more on the tertiary, or dispatch, level. Then she will ensure the control solution manages various operational objectives such as economical and energy-efficient microgrid power dispatch and pump schedule for water distribution, as well as preventing sudden changes in power output and pump’s flow rates to extend equipment lifespan.
She’s also focusing on resilience-informed control for the water-energy systems, using model-based predictive control strategy. To ensure resilient operation, she designs the control strategy to operate the water distribution pumps at their best efficiency point based on the pump’s performance curve combined with cost-effective operation based on hourly electricity cost. Additionally, she modifies the control strategy to consider ramp-up and ramp-down costs when providing the optimal solution of the microgrid’s power dispatch, thus avoiding unnecessary generator shutdowns or sudden output changes. The energy storage system is managed to balance the load and ensures it reaches full charge by the end of each day. Putri integrates resilience into the objective function of the control approach, so that the generated control solution will always focus on resilient operation—customers shouldn’t experience outages due to equipment that’s been damaged by the stress of demand.
The goal, Putri says, is to “make sure the consumer is always receiving water and electricity as required while enhancing the resilience of the instruments themselves. For example, if power demand suddenly increases, building resilience involves gradually ramping up the power dispatch to prevent damage to the instruments.”
For the rest of the week (Wednesday to Friday), Putri works with Khazaei and Moazeni on a Department Defense (DoD) Office of Naval Research-funded project to modernize the Navy’s microgrids. The research focuses on optimizing microgrids on secondary and primary levels in real time using a deep learning-based approach. Putri is concentrating on “developing the control strategies, making sure that the bus voltage [roughly, the total voltage in the system] is consistent at the nominal level and that the power sharing between the generation units satisfy the design requirements.”
One of the challenges she faces in developing these controls is the dual nature of the shipboard power systems: one system has a fairly steady energy consumption as it maintains the constant power of the ship’s propulsion and constant energy needs such as temperature control. The other system, Putri says, involves “pulse power loads, which features high frequency and intermittent loads,” providing power to “electromagnetic radar gun during missions.” These pulses can hit 5 to 10 megawatts and last for a few seconds. As a result, the control system faces the challenging task of balancing the ship’s infrastructure needs while delivering short, varied bursts of power that perform as needed in all situations. These are the controls Putri is working on; having previously designed model predictive control, she’s now working on data-driven controls to tackle these challenges. The overall goal for the shipboard power system is, as Putri says, “to satisfy load requirements, ensure safety and reliability of instruments and maintain voltages within the nominal requirements.”
Putri has been able to communicate her research and work collaboratively with her fellow researchers on writing and publishing articles and by presenting research at conferences. She has won several awards for her conference presentations. It’s impressive, Moazeni says: “She is in her third year of PhD and has already published multiple papers in high-end journals on topics for which she had zero background before she started her PhD.”
These activities also allow her to enhance her research and thinking about her work, as most articles go through several rounds of revisions. In general, Moazeni and Khazaei suggest revisions before the article is submitted, as do the journal editors and conference moderators during the submission and editing stages. Putri finds this helpful, especially as she’s from Indonesia and English is her second language. She also enjoys networking at conferences, because she says, she “gets to see all the people from the journals you read! And I can directly approach them and ask them how they do their research and receive ideas for how I can pursue my own research.” And that creates even more interconnections.
One example of her research and publishing an article published in the journal Applied Energy in October 2023. Putri and her coauthors, Moazeni and Khazaei, investigated the use of model predictive control to ensure demand compliance for integrated water-energy microgrids specified to islanded microgrids and water distribution system. Using data from Telford, Pennsylvania, the research used the interconnectedness of water and energy systems to regulate the use of mixed energy sources. This is done by penalizing the output power from generation units by their operation costs and the pump’s energy consumption by the hourly electricity cost. Their goal was to find out if the system could simultaneously satisfy both demand from water and energy systems while saving money and assess if it would benefit islanded or remote communities, which often have limited access to an electrical grid. Using the data, Putri found that the setup would satisfy the water demand with less energy use, and that it could save that town about $360,000 per year.
After she gets her degree, expected in June 2026, Putri would like to work for a national lab such as Argonne in Illinois, or Pacific Northwest National Laboratory in Washington and continue her research. She is open to most options, she says, but her biggest goal is to continue her research and make an impact in the water-energy realm.
Overall, she hopes her research and its applications can enhance the current practices of real-world systems by being directly implemented. Model predictive control, which Putri uses, considers system level control that combines both critical infrastructures. “PI [proportional-integral] control is widely used because it's straightforward and performs well,” she says. “But now, considering increased climate variation, we have more to consider when it comes to resiliency for all instruments. With model predictive control, we can address all these factors simultaneously and generate optimal control solutions. I’d like to develop an advanced control system yet remains user-friendly for a direct application to real world systems.”