Planning

PyEPlan provides a unified Investment and Operation Planning module (inosys) that addresses both long-term capacity expansion decisions and short-term dispatch optimization in a single optimization framework using mixed-integer linear programming (MILP).

Unified Investment and Operation Planning

The inosys module combines investment and operation planning in a single optimization problem, allowing for comprehensive microgrid analysis that considers both long-term investment decisions and short-term operational constraints simultaneously.

Mathematical Formulation

The unified planning problem is formulated as a mixed-integer linear programming (MILP) model:

Objective Function:

Minimize total system cost = Investment costs + Operation costs + Shedding costs

Where:

Investment costs = Capital costs for generators, storage, and renewables
Operation costs = Fuel costs + Variable O&M costs + Grid exchange costs
Shedding costs = Demand shedding + Renewable curtailment costs

Constraints:

Power Balance: Active and reactive power balance at each bus
Generator Constraints: Capacity limits, ramping constraints for conventional generators
Renewable Constraints: Maximum available power from solar and wind
Storage Constraints: Charge/discharge limits, state of charge constraints
Network Constraints: Line capacity limits, voltage limits
Investment Constraints: Binary variables for technology selection
Operational Constraints: Minimum up/down times, start-up costs

Key Features

Technology Selection: Optimal choice of renewable generators, energy storage, and conventional units
Capacity Sizing: Determination of optimal installed capacity for each technology
Geographic Siting: Optimal placement of generation and storage facilities
Economic Dispatch: Optimal power generation scheduling
Storage Management: Optimal charge/discharge scheduling with state of charge tracking
Demand Management: Load shedding and renewable curtailment optimization
Network Integration: Power flow and voltage constraints
Scenario Analysis: Handling of uncertainty through representative days

Usage Example

from pyeplan import inosys

# Initialize investment and operation planning system
planning_system = inosys(
    inp_folder="input_folder",
    ref_bus=0,
    dshed_cost=1000000,  # High cost to discourage load shedding
    rshed_cost=500,      # Cost for renewable curtailment
    phase=3,             # Three-phase system
    vmin=0.85,           # Minimum voltage limit
    vmax=1.15,           # Maximum voltage limit
    sbase=1,             # Base apparent power
    sc_fa=1              # Scaling factor
)

# Solve the unified optimization problem
planning_system.solve(
    solver='glpk',       # Optimization solver
    neos=False,          # Not using NEOS server
    invest=True,         # Include investment decisions
    onlyopr=False,       # Not operation-only mode
    commit=False,        # Not committing to NEOS
    solemail='',         # Solver email (for NEOS)
    verbose=False        # Verbose output
)

# Get optimization results
costs = planning_system.resCost()
wind_results = planning_system.resWind()
battery_results = planning_system.resBat()
solar_results = planning_system.resSolar()
conventional_results = planning_system.resConv()
curtailment_results = planning_system.resCurt()

Supported Technologies

Conventional Generators:

Diesel generators
Gas turbines
Combined heat and power (CHP) units
Grid connection (import/export)

Renewable Energy Sources:

Solar photovoltaic (PV) systems
Wind turbines
Hybrid renewable systems

Energy Storage:

Battery energy storage systems (BESS)
Pumped hydro storage
Thermal storage

Network Components:

Distribution lines and cables
Transformers
Switchgear and protection devices

Optimization Modes

Investment and Operation Mode (invest=True, onlyopr=False):

Optimizes both investment decisions and operational dispatch
Determines optimal technology mix and sizing
Provides comprehensive cost analysis

Operation-Only Mode (invest=False, onlyopr=True):

Optimizes only operational dispatch for existing infrastructure
Useful for operational analysis and cost assessment
Faster computation for large systems

Investment-Only Mode (invest=True, onlyopr=False with simplified operational constraints):

Focuses on long-term investment decisions
Uses simplified operational representation
Suitable for strategic planning

Solver Options

PyEPlan supports multiple optimization solvers through Pyomo:

Open-Source Solvers:

GLPK (GNU Linear Programming Kit) - Default
CBC (COIN-OR Branch and Cut)
IPOPT (Interior Point Optimizer)

Commercial Solvers:

Gurobi
CPLEX
MOSEK

Solver Selection Guidelines:

GLPK: Good for small to medium problems
CBC: Better for larger MILP problems
Gurobi/CPLEX: Best performance for large-scale problems
IPOPT: Suitable for continuous optimization problems

Results Analysis

The optimization results provide comprehensive information about:

Cost Analysis:

Total system cost breakdown
Investment costs by technology
Operational costs by component
Levelized cost of energy (LCOE)

Technology Mix:

Optimal installed capacity
Technology selection decisions
Geographic distribution

Operational Performance:

Hourly dispatch schedules
Storage state of charge profiles
Network power flows
Voltage profiles

Reliability Metrics:

Loss of load probability
Energy not served
System adequacy indicators

Integration with Other Modules

The planning module integrates seamlessly with other PyEPlan modules:

Data Processing Integration:

Uses representative days from datsys module
Incorporates renewable generation profiles
Handles load demand scenarios

Network Integration:

Incorporates network topology from rousys module
Considers line parameters and constraints
Optimizes power flow distribution

This integrated approach ensures that all aspects of microgrid planning are considered in a unified optimization framework, leading to more robust and cost-effective solutions.