Progress Utilities Module

class progress.mod_utilities.RAUtilities

Bases: object

Contains utility methods required for performing mixed time sequential Monte Carlo (MCS) simulations and evaluating reliability indices in power systems.

This class offers functions to:

  • Compute reliability transition rates for various components (generators, transmission lines, ESS).

  • Track and update component states (fail/repair) over simulated time.

  • Aggregate or compute net capacities, state of charge, and wind/solar generation.

  • Perform optimization-based dispatch (economic dispatch) in either a zonal or copper-sheet model.

  • Compute final reliability indices (e.g., LOLP, EUE, LOLE) and handle parallel (MPI) processing steps.

CheckConvergence(s, LOLP_rec, comm, rank, size, mLOLP_rec, COV_rec)

Gathers partial results from all MPI processes and updates the running mean and coefficient of variation (COV) for LOLP. Used to track convergence during MCS.

Parameters:

  • s (int) – Current sample index.

  • LOLP_rec (np.ndarray) – Array of LOLP values recorded by the local process for each sample so far.

  • comm (mpi4py.MPI.Comm) – MPI communicator object.

  • rank (int) – MPI rank of this process.

  • size (int) – Total number of MPI processes.

  • mLOLP_rec (np.ndarray) – Array for storing the running mean of LOLP across samples.

  • COV_rec (np.ndarray) – Array for storing the coefficient of variation (COV) at each sample.

Returns:

A 2-element tuple containing:

  • mLOLP_rec (np.ndarray): Updated array of mean LOLP values across processes.

  • COV_rec (np.ndarray): Updated array of COV values, checking for convergence.

Return type:

tuple

GetReliabilityIndices(indices_rec, sim_hours, samples)

Computes final reliability indices (LOLP, EUE, EPNS, LOLE, etc.) by averaging over the recorded arrays.

Parameters:

  • indices_rec (dict) – Dictionary with arrays of reliability indices (one entry per sample).

  • sim_hours (int) – Total number of hours (e.g., 8760) in the simulation horizon.

  • samples (int) – Number of Monte Carlo samples performed.

Returns:

A dictionary containing final reliability metrics such as LOLP, LOLH, EUE, EPNS, LOLE, etc.

Return type:

dict

NextState(t_min, ng, ness, nl, lambda_tot, mu_tot, current_state, cap_max, cap_min, ess_units)

Determines which component (gen/trans/ESS) transitions (fails or repairs) next using exponential random draws, then updates system states and capacities.

Parameters:

  • t_min (float) – Time (hours) remaining until the next state transition event occurs.

  • ng (int) – Number of conventional generators.

  • ness (int) – Number of energy storage systems.

  • nl (int) – Number of transmission lines.

  • lambda_tot (np.ndarray) – Failure rates for all components (ordered: gen, lines, ESS).

  • mu_tot (np.ndarray) – Repair rates for all components (ordered: gen, lines, ESS).

  • current_state (np.ndarray) – Current binary states of each component (1 = up, 0 = down). For ESS, each integer denotes how many are up among identical units.

  • cap_max (np.ndarray) – Maximum capacities for each component type.

  • cap_min (np.ndarray) – Minimum capacities for each component type.

  • ess_units (np.ndarray) – Number of identical units for each ESS installation.

Returns:

A 3-element tuple containing:

  • current_state (np.ndarray): Updated states after one component fails/repairs.

  • current_cap (dict): Current maximum and minimum capacity arrays (accounting for the updated state).

  • t_min (float): Updated time to the next state transition.

Return type:

tuple

OptDispatch(ng, nz, nl, ness, fb_ess, fb_soc, BMva, fb_Pg, fb_flow, A_inc, gen_mat, curt_mat, ch_mat, gencost, net_load, SOC_old, ess_pmax, ess_eff, disch_cost, ch_cost)

Formulates and solves an economic dispatch (transportation model) optimization for a single hour, including line flow constraints, generation dispatch, storage charging/discharging, and load curtailment.

Parameters:

  • ng (int) – Number of conventional generators.

  • nz (int) – Number of zones.

  • nl (int) – Number of lines.

  • ness (int) – Number of energy storage systems.

  • fb_ess (function) – Function specifying variable bounds for ESS charging power.

  • fb_soc (function) – Function specifying variable bounds for ESS state of charge.

  • BMva (float) – System base power (MVA).

  • fb_Pg (function) – Function specifying variable bounds for generator and ESS discharge power.

  • fb_flow (function) – Function specifying variable bounds for line flow.

  • A_inc (np.ndarray) – Incidence matrix for lines.

  • gen_mat (np.ndarray) – Generation matrix indicating which bus/zone each generator belongs to.

  • curt_mat (np.ndarray) – Matrix used for load curtailment variables (identity mapping to each zone).

  • ch_mat (np.ndarray) – Matrix used for storage charging variables (which bus/zone each ESS belongs to).

  • gencost (np.ndarray) – Per-unit fuel costs for conventional generators.

  • net_load (np.ndarray) – Net load array (MW) per zone.

  • SOC_old (np.ndarray) – Previous hour’s state-of-charge for each ESS (MWh).

  • ess_pmax (np.ndarray) – Maximum power output (MW) for each ESS.

  • ess_eff (np.ndarray) – Round-trip efficiency (fraction) for each ESS.

  • disch_cost (np.ndarray) – Discharge cost for each ESS.

  • ch_cost (np.ndarray) – Charge cost for each ESS.

Returns:

A 2-element tuple containing:

  • load_curt (float): Total load curtailment (MW) in this hour.

  • SOC_old (np.ndarray): Updated state-of-charge for each ESS after dispatch.

Return type:

tuple

OptDispatchLite(ng, nz, ness, fb_ess, fb_soc, BMva, fb_Pg, A_inc, gencost, net_load, SOC_old, ess_pmax, ess_eff, disch_cost, ch_cost)

Solves an economic dispatch problem ignoring line flow constraints (copper-sheet model). This method is a simplified version of OptDispatch().

Parameters:

  • ng (int) – Number of conventional generators.

  • nz (int) – Number of zones.

  • ness (int) – Number of energy storage systems.

  • fb_ess (function) – Function specifying variable bounds for ESS charging power.

  • fb_soc (function) – Function specifying variable bounds for ESS state-of-charge.

  • BMva (float) – System base power (MVA).

  • fb_Pg (function) – Function specifying variable bounds for generator and ESS discharge power.

  • A_inc (np.ndarray) – Incidence matrix (not strictly used, but included for consistency).

  • gencost (np.ndarray) – Per-unit costs for conventional generators.

  • net_load (np.ndarray) – Net load (MW) for the entire system (summed over all zones).

  • SOC_old (np.ndarray) – Previous hour’s state-of-charge for each ESS (MWh).

  • ess_pmax (np.ndarray) – Maximum power output for each ESS.

  • ess_eff (np.ndarray) – Round-trip efficiency for each ESS.

  • disch_cost (np.ndarray) – Discharge cost for each ESS.

  • ch_cost (np.ndarray) – Charge cost for each ESS.

Returns:

A 2-element tuple containing:

  • load_curt (float): Total load curtailment (MW).

  • SOC_old (np.ndarray): Updated state-of-charge for each ESS.

Return type:

tuple

OutageAnalysis(var_s)

Identifies the start and end of each outage (load-curtailed period) from the binary LOL array and calculates their durations.

Parameters:

var_s (dict) – Dictionary of sample-specific tracking variables (e.g., “label_LOLF”).

Returns:

1D array of outage durations (in hours).

Return type:

np.ndarray

OutageHeatMap(LOL_track, size, samples, main_folder)

Aggregates hourly loss-of-load data into a monthly/hourly matrix and writes the percentage of load-loss hours by month/hour to a CSV file. Can be used later for creating heatmaps.

Parameters:

  • LOL_track (np.ndarray) – 4D array of shape (size, samples, 365, 24) representing the LOL states.

  • size (int) – Total number of MPI processes.

  • samples (int) – Number of Monte Carlo samples.

  • main_folder (str) – Directory path where results (CSV/plots) are saved.

Returns:

None

Return type:

None

ParallelProcessing(indices, LOL_track, comm, rank, size, samples, sim_hours)

Gathers reliability indices and the LOL tracker from all MPI processes to compute average indices over all ranks. Optionally saves final results and calls OutageHeatMap() if a full-year simulation is run.

Parameters:

  • indices (dict) – Dictionary of reliability indices for the local rank.

  • LOL_track (np.ndarray) – 2D array of shape (samples, hours) tracking LOL states for the local rank.

  • comm (mpi4py.MPI.Comm) – MPI communicator object.

  • rank (int) – MPI rank of this process.

  • size (int) – Total number of MPI processes.

  • samples (int) – Number of Monte Carlo samples performed.

  • sim_hours (int) – Number of hours in each sample (e.g., 8760).

Returns:

If rank=0, a dictionary containing aggregated reliability indices (LOLP, EUE, etc.) across all ranks, else returns None.

Return type:

dict or None

SolarPower(n, nz, s_zone_no, solar_prob, s_profiles, s_sites, s_max)

Determines solar power output for each zone based on a randomly selected cluster/day/hour profile.

Parameters:

  • n (int) – Current hour in the overall MCS simulation.

  • nz (int) – Number of zones.

  • s_zone_no (np.ndarray) – Zone indices for each solar site.

  • solar_prob (np.ndarray) – Array of cluster probabilities (rows = clusters, columns = months).

  • s_profiles (list[np.ndarray]) – List of 3D arrays, each containing daily solar profiles for a cluster (dimensions: [days, 24 hours, sites]).

  • s_sites (int) – Number of solar sites.

  • s_max (np.ndarray) – Maximum capacity (MW) of each solar site.

Returns:

A 2D NumPy array of shape (24, nz), giving solar generation (MW) for each zone over 24 hours. Only the row corresponding to hour n%24 is relevant at time n, but the entire 2D block is returned for convenience.

Return type:

np.ndarray

TrackLOLStates(load_curt, BMva, var_s, LOL_track, s, n)

Records whether load was curtailed at the current hour, and logs loss-of-load days, frequencies, etc.

Parameters:

  • load_curt (float) – The amount of load curtailment (MW).

  • BMva (float) – System base power (MVA) for consistent unit conversions.

  • var_s (dict) – Dictionary of temporary simulation variables (e.g., curtailment array, LOL day counts).

  • LOL_track (np.ndarray) – 2D array (shape: [samples, hours]) tracking which hours had loss of load.

  • s (int) – Current Monte Carlo sample index.

  • n (int) – Current hour index.

Returns:

A 2-element tuple containing:

  • var_s (dict): Updated dictionary of simulation variables.

  • LOL_track (np.ndarray): Updated array marking loss-of-load hours.

Return type:

tuple

UpdateIndexArrays(indices_rec, var_s, sim_hours, s)

Updates reliability index accumulators (LOLP, EUE, EPNS, LOLE, etc.) after each sample is completed.

Parameters:

  • indices_rec (dict) – Dictionary that stores reliability indices across samples.

  • var_s (dict) – Dictionary of sample-specific tracking variables (curtailment array, LOL states, etc.).

  • sim_hours (int) – Number of hours in the current simulation (e.g., 8760 for a year).

  • s (int) – Current sample index.

Returns:

indices_rec (dict): Updated dictionary containing reliability index arrays for all samples.

Return type:

dict

WindPower(nz, w_sites, zone_no, w_classes, r_cap, current_w_class, tr_mats, p_class, w_turbines, out_curve2, out_curve3)

Computes wind power generation by randomly transitioning wind speed classes, then mapping each class to a turbine power output.

Parameters:

  • nz (int) – Number of zones in the system.

  • w_sites (int) – Number of wind farm sites.

  • zone_no (np.ndarray) – Zone indices for each wind site.

  • w_classes (int) – Total number of wind speed classes used.

  • r_cap (np.ndarray) – Rated capacities (MW) of wind turbines at each site.

  • current_w_class (np.ndarray) – Current wind speed class for each site (1D array).

  • tr_mats (np.ndarray) – 3D array of transition rate matrices for each site (dimensions: [site, w_classes, w_classes]).

  • p_class (np.ndarray) – Array specifying the power curve classification type (e.g., 2 or 3) for each site.

  • w_turbines (np.ndarray) – Number of turbines at each site.

  • out_curve2 (np.ndarray) – Power output curve for class 2 turbines (indexed by wind speed class).

  • out_curve3 (np.ndarray) – Power output curve for class 3 turbines (indexed by wind speed class).

Returns:

A 2-element tuple containing:

  • w_zones (np.ndarray): Wind generation in each zone (size = nz).

  • current_w_class (np.ndarray): Updated wind speed classes for the next hour.

Return type:

tuple

capacities(nl, pmax, pmin, ess_pmax, ess_pmin, cap_trans)

Constructs arrays of maximum and minimum capacities for generators, lines, and ESS for use in the MCS.

Parameters:

  • nl (int) – Number of transmission lines.

  • pmax (np.ndarray) – Maximum capacities (MW) of each generator.

  • pmin (np.ndarray) – Minimum capacities (MW) of each generator.

  • ess_pmax (np.ndarray) – Maximum power outputs (MW) for each ESS.

  • ess_pmin (np.ndarray) – Minimum power outputs (MW) for each ESS.

  • cap_trans (np.ndarray) – Transmission capacities (MW) for each line.

Returns:

A 2-element tuple containing:

  • cap_max (np.ndarray): Concatenated maximum capacities (generators + lines + ESS).

  • cap_min (np.ndarray): Concatenated minimum capacities (generators + lines + ESS).

Return type:

tuple

reltrates(MTTF_gen, MTTF_trans, MTTR_gen, MTTR_trans, MTTF_ess, MTTR_ess)

Computes repair and failure rates for generators, transmission lines, and ESS.

Parameters:

  • MTTF_gen (np.ndarray) – Mean time to failure for generators (hours).

  • MTTF_trans (np.ndarray) – Mean time to failure for transmission lines (hours).

  • MTTR_gen (np.ndarray) – Mean time to repair for generators (hours).

  • MTTR_trans (np.ndarray) – Mean time to repair for transmission lines (hours).

  • MTTF_ess (np.ndarray) – Mean time to failure for energy storage systems (hours).

  • MTTR_ess (np.ndarray) – Mean time to repair for energy storage systems (hours).

Returns:

A 2-element tuple containing:

  • mu_tot (np.ndarray): Repair rates (1/MTTR) for all components (gen, trans, ESS).

  • lambda_tot (np.ndarray): Failure rates (1/MTTF) for all components.

Return type:

tuple

updateSOC(ng, nl, current_cap, ess_pmax, ess_duration, ess_socmax, ess_socmin, SOC_old)

Adjusts the state-of-charge (SOC) for storage units to reflect any failures/repairs that reduce or restore capacity.

Parameters:

  • ng (int) – Number of generators.

  • nl (int) – Number of lines.

  • current_cap (dict) – Dictionary of current max and min capacities (from NextState()).

  • ess_pmax (np.ndarray) – Original maximum power outputs for ESS (MW).

  • ess_duration (np.ndarray) – Duration (hours) of each ESS at rated power.

  • ess_socmax (np.ndarray) – Maximum SOC fraction for each ESS.

  • ess_socmin (np.ndarray) – Minimum SOC fraction for each ESS.

  • SOC_old (np.ndarray) – Previous SOC values (in MWh).

Returns:

A 3-element tuple containing:

  • ess_smax (np.ndarray): Maximum allowable SOC (in MWh) after accounting for failures/repairs.

  • ess_smin (np.ndarray): Minimum allowable SOC (in MWh).

  • SOC_old (np.ndarray): Updated SOC for each ESS.

Return type:

tuple