The LGO program offers energy management internships, especially related to delivering and using electricity and gas. Other companies offer projects around sustainable energy usage, solar energy, and fuel cells. Internships provide a unique opportunity to pair a student’s engineering and MBA studies with an energy engineering and management concentration. For students looking to go even deeper, the Nuclear Science and EngineeringMechanical Engineering and Civil Engineering departments have specializations in energy, and MIT’s Energy Initiative is a campus-wide organization focused on the topic.

Modeling Heat Pump Adoption Propensity and Simulating the Distribution-Level Effects

Trevor Thompson (LGO ’21)

Engineering Department: Mechanical Engineering
Company: National Grid

Problem: To maintain alignment with regional carbon reduction goals, National Grid must find innovative ways to reduce greenhouse gas emissions in its service territories. For the heating sector in particular, air source heat pumps (ASHP) represent a promising opportunity to decarbonize the residential space conditioning market through electrification. To meet these goals, the company must accelerate the current rate of customer adoption – utilizing data and tools at its current disposal.

Trevor Thompson LGO '21 Thesis Diagram
The methods of propensity modeling can be utilized for any “product adoption” scenario across industries.

Approach: Trevor began with collecting relevant data to model the propensity of customers to adopt ASHPs. Sources of data included: customer account data, energy efficiency program participation data, electricity and natural gas infrastructure, customer demographics, housing data, and survey data. He then developed and deployed a machine learning pipeline to build predictive models using a variety of pre-processing, feature selection, hyperparameter, and training algorithm options.

Impact: Using the propensity modeling results to simulate distribution system impacts from ASHP adoption, Trevor was able to demonstrate a high-accuracy capability to “geotarget” customers. Understanding who the most likely customers are to adopt and customizing marketing efforts to target these customers will help to accelerate early adoption. This capability also allows for more granular planning capabilities with regards to zonal electrification and alleviating natural gas system supply constraints. This information can be used to better understand the location and timing of future distribution system upgrades.

Transmission System Overvoltage Mitigation Through the Use of Distributed Generation (DG) Smart Inverters

Ali Said Alrayes (LGO ’20)

Engineering Department: Electrical Engineering and Computer Science
Company: National Grid
Location: Waltham, MA

Alrayes, National Grid, 2020
Ali proved that implementing a smart inverter-based solution had significant potential cost savings for National Grid.

Problem: Ali’s project at National Grid focused on rapid deployment of Distributed Generation (DG) across its electricity distribution system. National Grid predicted that increased DG deployment had the potential to cause overvoltage violations on the transmission system, due to the underloading of high voltage (69kV and 115kV) transmission lines. Traditional responses were to install system upgrades in the form of shunt reactors. However, these system upgrades are expensive and time-consuming to install, potentially de-incentivizing and/or delaying the deployment of DG projects. 

Approach: Ali proposed utilizing the reactive power absorption capability of the DG inverters, in aggregation, to absorb excessive reactive power from the system making substations act as “Virtual” Shunt Reactors. To prove this, Ali investigated the maximum capability of a distribution feeders reactive power absorption through their DG smart inverters, by modeling two “representative” distribution feeders and substations under different deployment scenarios, labeled “Reactive Power Hosting Capacity. Ali then performed a cost-benefit analysis to compare the proposed smart-inverter solution to the existing shunt reactor solution to prove the viability of the project. 

Impact: Ali’s results proved that under specific scenarios, implementing a smart inverter-based solution had potential cost savings up to $300,000/year/substation compared to installing shunt reactors. As National Grid moves into a new era of greater collaboration with DG owners, Ali’s model offers a roadmap for future work that includes redesigning the interconnection standards in Massachusetts to require smart inverter installations on all DG sites, and studying additional feeders with a larger diversity of characteristics.