Data for optimization of Electric Vehicle charging configuration to reduce overloading in low voltage networks
Description
By optimizing for a reduction in peak line loading in a typical benchmark 0.4 kV network, additional line loading percentages caused by the introduction of incremental increases in the number of EVs charging in an agent based Monte Carlo simulation can be reduced by over 10% in the worst case and about 5% on average across several cable types. Optimization for a reduction in peak voltage drops shows an increase in voltage available at different nodes by up to 7 V in the worst case and 1.5 V on average. Optimization for total line losses shows a negligible savings across low voltage networks
Files
Steps to reproduce
The files titled combined_(X)cars(Y).csv (where X is number of cars in simulation and Y is hour) are sampled from a normal distribution of individual household electricity consumption for columns 1-11. Columns 12 to22 are sampled from lognormal distribution representing probability of cars charging at that hour given Y cars in the total simulation. The filename series upd_(A)_limit(B)readjusted(C).csv represents results from simulation after optimizing EV charging configuration. A represents the optimization variable (load: peak line loading, loss: total cable losses and volts: voltage drops). B is either 0: no limit on charging configuration or 1: no charging node can charge more than 25% of the total cars charging at that hour and C represents the total number of cars in the simulation.