Skip to content
Back to Projects

Gravity-Buoy Energy Storage System

September 1, 2023Completedacademic
matlab modeling

Overview

In this group project at NTNU (Norwegian University of Science and Technology), our team of 6 modeled and simulated a gravity-buoy energy storage system in MATLAB/Simulink. The concept: store excess wind energy from turbines in the Flevopolder (Netherlands) by submerging buoyant spheres underwater, then release them to generate electricity for office buildings in Amsterdam.

Full Report — Gravity-Buoy Energy Storage

Concept

The system works on a simple principle:

  1. Charging — When wind turbines produce excess energy, it powers winches that pull buoyant spheres down to the seabed (1 km depth), storing gravitational potential energy
  2. Discharging — When energy is needed, the spheres are released and their buoyancy drives generators as they rise, producing electricity
  3. Transmission — Power is transmitted to Amsterdam office buildings via cables, with Ohmic losses modeled

The net buoyancy force driving power generation:

F=(V(ρseawaterρhydrogen)mcable)gF = (V \cdot (\rho_{seawater} - \rho_{hydrogen}) - m_{cable}) \cdot g

Simulink Model

The model incorporates:

  • Buoyancy physics — Archimedes' principle with seawater density
  • Drag forces — Hydrodynamic resistance during ascent/descent
  • Cable mechanics — Mass and elasticity of tethering cables
  • Power transmission losses — Ohmic losses over the transmission distance
  • Energy balancing — Monthly wind generation vs. building consumption profiles

Results

The simulation showed that approximately 40% of the building energy demand had to be bought externally — the gravity-buoy system could cover about 60% of needs. Key losses came from:

  • Hydrodynamic drag during sphere ascent
  • Cable weight reducing net buoyancy
  • Transmission line losses

The system works in principle but has significant round-trip efficiency limitations compared to more established storage technologies.

Results & Discussion

This project was done during an exchange period at NTNU, which gave me exposure to a different academic environment and team dynamics. The Simulink modeling was the most technically valuable part — building a physical system model from first principles (buoyancy, drag, power transmission) and validating it against energy balance requirements. The honest result — that the system underperforms — was itself a useful engineering conclusion.

Technologies Used

MATLAB, Simulink, energy system modeling, buoyancy physics, LaTeX