User Guide

Design Problem Description

SHIPP relies on two classes to describe a sizing optimization problem:

• a Production object describe a renewable energy production (e.g. wind or solar PV) through

  • its power production in time power. Note that in the optimization problem, the power can be curtailled.

  • its cost per power capacity p_cost

• a Storage object describe abstract storage systems through:

  • its cost per energy and power capacity e_cost and p_cost

  • its energy and power capacity e_cap and p_cap

  • its efficiency in charge and discharge eff_in and eff_out. Note that only constant efficiencies are considered here.

Finally, the optimization problem requires the time series of electricity prices on the day-ahead market.

Solving the optimization problem

The optimization problem is formulated and solved using a command of the form:

os = solve_lp_pyomo(price, production1, production1, storage1, storage2, discount_rate, n_year, p_min, p_max, n, name_solver)

where the discount rate and the number of years n_year are used for the calculation of the NPV. The parameters p_min and p_max are used to describe the constraints for the maximum and minimum power production. The number of time steps is n. The parameter name_solver refers to a solver compatible with pyomo, for example ‘mosek’, ‘cplex’, ‘gurobi’.

Alternatively, the problem can be solved with the built-in solver in scipy, scipy.optimize.linprog with the following command:

os = solve_lp_sparse(price, production1, production1, storage1, storage2, discount_rate, n_year, p_min, p_max, n)

Results post-processing

The optimization results are stored in a OpSchedule object describing the operation schedule of the power plant, with the following members:

• production_list: a list of Production objects corresponding to the input objects of the optimization problem

• storage_list: a list of Storage objects corresponding to the input objects of the optimization problem. However, their power and energy capacity correspond to the optimal design.

• production_p: a list of the power production for the objects in production_list. In case of curtailment, the first production unit assumes all the curtailment and the second production unit operates at full power.

• storage_p: list of optimal power operation (charge/discharge) for the storage objects

• storage_e: list of optimal energy level evolution for the storage objects

• losses: list of power losses corresponding to the storage objects.

• power_out: total power to the grid for the production and storage units

• revenue: total annual revenues from selling electricity.

• revenue_storage: revenue contribution from the storage units.

• npv: NPV for the total system

• a_npv: Added NPV due to the addition of storage units

• irr: Internal Rate of Return for the power plant.

The operation schedule can be visualized using the following commands:

• os.plot_powerflow(): line plot of the power production and the energy level evolution

• os.plot_powerout(xlim = [start_time, end_time]): bar plot of the power send to the grid