Department: Structural Engineering
Faculty Advisor(s): Michael D. Todd

Primary Student
Name: Eric John Kjolsing
Email: ekjolsin@ucsd.edu
Phone: 858-534-5993
Grad Year: 2016

A vibration based energy harvesting system is desired to replace or supplement existing components used to power downhole well monitoring equipment. Since the geometry, loads, and boundary conditions vary from well to well, the energy harvesting system must be well-specific in order to maximize the harvested energy. This poster describes a MATLAB program written to facilitate the design of each well-specific energy harvester. The program consists of three sequential design phases. In the first phase, user-defined representative acceleration profiles are transformed into the Fourier domain to generate a time-varying 360 frequency profile. After accounting for geometric limitations within the energy harvesting system, this frequency profile is used to calculate signal power estimates for all relevant permutations of the base frequencies of the individual energy harvester elements. Each of these realizations is then integrated over the time domain of interest and compared to all other realizations in order to determine the optimal base frequencies for each energy harvesting element within the energy harvesting system. In the second phase, a simple damping model is used to generate a mass normalized power profile to give the user an order of magnitude estimate of the power that might be extracted by the energy harvesting system. If the power estimate is significantly below the power required for the well-specific monitoring equipment then the energy harvester design can be abandoned before phase three; phase three is significantly more input and resource intensive than the first two phases. In the third phase, uncertainties in the mechanical and electrical properties are accounted for. User-defined properties are made to be time and temperature dependent and given bounds of uncertainty. Power estimates are then generated using the coupled equations of motion (mechanical and electrical) in a Monte Carlo simulation. The output realizations are condensed into an expected power profile which can be used in determining which well monitoring equipment can be partially or fully powered by the energy harvesting system. Power profiles based on confidence intervals are also provided. The described program will be of use to well operators who may be thinking of deploying an energy harvesting system and engineers/researchers interested in designing and fabricating downhole energy harvesting systems.

Industry Application Area(s)
Civil/Structural Engineering | Energy/Clean technology | Hydrocarbon

« Back to Posters or Search Results

Contact:   researchexpo@soe.ucsd.edu   (858) 534-6068