Sumanik, Spencer R. D.

Coupling energy storage systems with renewable generation systems has a variety of advantages including: increasing the penetration of renewables, reducing curtailment, providing energy security, lowering greenhouse gas emissions and providing an array of ancillary services. These advantages are multiplied for hybrid off-grid power systems. However, sizing both the energy storage system for both energy capacity and the power rating can be difficult in these remote locations, as modularity, solution characteristics, transportation and installation logistics must be taken into consideration. This analysis simulates one year of energy generation and load based on real data from the remote community of Old Crow Yukon Territory, Canada. Throughout the year data pertaining to the effect of the energy storage system on the remote power system are collected. This information is used to find the optimal energy capacity and power rating for the remote isolated power system, using a specific energy storage dispatch logic. Seven parameters are used to find the optimal solution: cost of energy, cost of power, cost of fuel, cost of curtailment, cost of reserve, cost of insufficient power, and cost of blackouts. Each parameter is compared on the basis of levelized cost of energy and a specific weighting factor. Three cases are examined for different weighing factors, to demonstrate the flexibility of the approach to serve specific purposes. Case 1 examines equal values across all weighting factors. Case 2 focuses on fuel consumption and curtailment and Case 3 focuses on energy security and reliability.

This study investigates a general model for fuel consumption of prime diesel generators within a set range of loadings. The model is parametrized by the rated power output, or sizing, of a generator. This research gives insight into generator fuel consumption characteristics across a wider population of generators to provide a general estimate of fuel consumption for a given sizing. Manufacturer data sheets containing fuel consumption measurements are collected online and through electric power utilities in the Canadian territories for 40 unique diesel generator sizings. The sizing group effects are accounted for through a non-pooled and multilevel regression. Subsequent estimates of linear parameters across generator sizings are modelled through ordinary least squares to obtain the desired model for fuel consumption. The generality and adequacy of this model is investigated through simulation and selected fresh data sources. The general model for prime diesel generator fuel consumption serves as a useful estimate or approximation for subsequent work that requires a general fuel efficiency estimate.