• Yet Another Power System Model (YAPOS)
  • Installation
  • Getting started
  • Launching a simulation
  • Notes and caveats
  • How to set up and launch a simulation
  • Formulation
  • Inputs
  • Outputs
  • Available simulations

What is YAPOS? YAPOS is an economic dispatch model implemented in Python using Pyomo. The problem is defined at daily time-scale and it aims to provide a simple, yet insightful, modelisation of national power systems.

Why another power system model? Although several sophisticated power system model already exist with open license (for example Dispa-SET, PyPSA), YAPOS provides an alternative to all the users that:

1. Do not want or cannot use a commercial solver (e.g. GUROBI or CPLEX)
2. Want something light to run (a simulation for 34 countries run in a couple of minutes)
3. Need a tool able to run many simulations in batch in a tidy way

How to use YAPOS? YAPOS has been developed and designed aiming to build an impact model for all the people investigating the link between energy & meteorology. However, it can be used whenever you need a lightweight power system model able to reproduce the impact of meteorological factors.

YAPOS consists of a single Python file which requires the following modules: - pyomo - numpy - pandas - xarray with netcdf4

You can install manually the modules with pip or other package managers. If you use anaconda you can easily create an environment for YAPOS with the following command: conda env create -f environment.yml

From you Python environment you can run a test simulation with the command: python main.py db/test/ test_run

The output of the simulation can be found in the folder db/test.

With YAPOS code you can also find a database of simulations already available. More information here.

To launch a simulation you need to execute the script main.py specifying two arguments: the path of the folder containing the input files and the name of the simulation. This name will be used as name of the NetCDF output.

• The model uses Cbc solver as default solver, however the code might be easily adapted to use other solvers (e.g. GLPK)
• The model interprets the input files in a “positional” fashion because our formulation converts the CSVs into information by position (using Pandas iloc). Then be careful when changing the inputs, all the dimensions (units, zones, etc.) MUST follow the same order in all the input files and the name of the simulation that will be used to name the NetCDF output file.

1. Create a folder for each simulation
2. Copy inside the folder the input files needed for the simulation

• Generation units: gen.csv
• Transmission lines: lin.csv
• Demand time-series: dem.csv
• Renewable time-series: ren.csv
• Renewables (non-dispatchable) generation capacity: ren_pp.csv
• Inflow time-series: inflow.csv
• Availability time-series: avail.csv
• Minimum storage levels: stomin.csv

3. Run the model with Python

A companion package written in R named yaposer is available in [its repository](https://github.com/matteodefelice/yaposer. The package helps the user to load, save, edit and visualisize simulation inputs and results.

Read this page for a tutorial

YAPOS uses a linear formulation at daily resolution. There are seven continuous variables (and three slack variables) defined over four sets.

Sets

Optimization variables

Slack variables

The optimization problem uses the following objective function:

The formulation uses 13 different constraints:

• Electricity balance
• Minimum generation for the units
• Maximum generation for the units
• Maximum rate of ramping-up for hydropower
• Maximum rate of ramping-down for hydropower
• Minimum storage level
• Maximum storage level
• Storage balance
• Maximum discharge for hydropower
• Minimum power flow
• Maximum power flow

As explained before YAPOS needs seven input files.

File name is dem.csv and contains a time-series per zone of the electricity demand.

Daily demand

The file has the name gen.csv and contains a table with a row per each dispatchable generating unit (i.e. no variable generation units like wind and solar). Each row has the following fields:

• Unit: unit name
• bus: the index (from 0 to N) of the zone where the unit belongs. The zones are in the order specified in the demand files (and also non-dispatchable renewables)
• Technology and Fuel: code of technology and fuel based on the Dispa-SET classification
• cost: cost of electricity generation (EUR/MWh)
• co2_per_mwh: CO2 emission per MWh
• max: Maximum daily generation of the unit (MWh)
• stomax: Maximum value of the storage (MWh)
• min: Minimum daily generation of the unit (MWh)
• stomin: Minimum value of the storage (MWh)

Here an example of the file db/envarclim/1990/gen.csv.

Generation units

The file has the name lin.csv and contains a table with a row per each line. Lines represents the cross-border transmission lines connecting two zones. Each line consists of the name of the line, the two connected zones (used an integer index as in the bus field in the generation units) and the maximum daily capacity (MWh).

Here an example.

Transmission lines

File name is ren.csv and contains a time-series per zone of the generation from non-dispatchable renewables. The amount of electricity specified will be generated unless curtailed.

Daily renewable generation

The file has the name ren_pp.csv and contains a table with a row per each non-dispatchable fuel/technology (variable generation units like wind and solar). Each row has the following fields:

• Technology: code of technology based on the Dispa-SET classification
• The capacity in MW for each zone

NOTE This file is not actually used for the simulation but it might be useful to the modeller to understand the energy mix “contained” into the non-dispatchable electricity time-series

Here an example of the file db/envarclim/1990/ren_pp.csv.

Non-dispatchable generation capacity

File name is inflow.csv and contains a time-series per unit of the available energy available (MWh) for hydropower generation each day. The file contains a column for all the units, then non-hydro units will have a constant zero value.

Daily hydropower inflow

File avail.csv contains a time-series per zone of the availability factor, a value between 1 (100% available generation) and 0 (unit non available) for each unit.

Daily unit availability

This file, named stomin.csv, contains a time-series per unit of the minimum level of the storage. The value will be between stomin and stomax specified in the list of generation units. As for the inflow the table has a column for each unit, with a constant zero for all the units without storage.

Daily minimum storage

At the end of a YAPOS simulation, in the same folder of the input files the model will save a set of files containing the results for the simulation.

The main output file is in NetCDF format and its filename is set with the keyword specified when executing the simulation. In addition to the NetCDF file, the simulation saves also a set of CSV files in addition to a file named model.lp, containing the linear programming formulation (that can be used for example to rerun the same problem with another solver, e.g. GLPK).

import xarray as xr
xr.open_dataset('/Users/matteodefelice/work/yaposer/inst/extdata/es-pt-fr.nc')

## <xarray.Dataset>
## Dimensions:            (day: 365, line: 2, unit: 26, zone: 3)
## Coordinates:
##   * day                (day) int64 0 1 2 3 4 5 6 ... 358 359 360 361 362 363 364
##   * unit               (unit) object 'FR_Hydro reservoir' ... 'PT_Coal fleet Supercritical'
##   * zone               (zone) object 'ES' 'PT' 'FR'
##   * line               (line) object 'ES-FR' 'ES-PT'
## Data variables:
##     production         (day, unit) float64 ...
##     availability       (day, unit) float64 ...
##     inflow             (day, unit) float64 ...
##     storage_min        (day, unit) float64 ...
##     demand             (day, zone) int64 ...
##     renewables         (day, zone) float64 ...
##     flow               (day, line) float64 ...
##     shed_load          (day, zone) float64 ...
##     curtailed          (day, zone) float64 ...
##     storage_level      (day, unit) float64 ...
##     water_slack        (day, unit) float64 ...
##     storage_slack      (day, unit) float64 ...
##     curtailment_slack  (day, zone) float64 ...
##     unit_bus           (unit) int64 ...
##     unit_Technology    (unit) object ...
##     unit_Fuel          (unit) object ...
##     unit_cost          (unit) float64 ...
##     unit_co2_per_mwh   (unit) float64 ...
##     unit_max           (unit) float64 ...
##     unit_stomax        (unit) int64 ...
##     unit_min           (unit) float64 ...
##     unit_stomin        (unit) int64 ...
##     line_from          (line) int64 ...
##     line_to            (line) int64 ...
##     line_cap           (line) int64 ...
## Attributes:
##     created:          20/05/2020 21:44:19
##     hostname:         New-MacBook-Pro-2.local
##     model_folder:     db/test/
##     simulation_name:  test_run

YAPOS model writes the following files at end the of the simulation:

• prod.csv: daily generation for each generation unit
• flow.csv: electricity flow for each transmission line, positive signs indicate a flow from the first zone to the second one specified in the line name (e.g. CH-IT)
• shed.csv: shed load for each zone
• curt.csv: curtailed electricity for each simulated zone
• ws.csv: value of the water slack variable for each generation unit. This is the slack variable for the storage balance constraint
• ss.sv: value of the storage slack variable for each generation unit. This is the slack variable for the minimum storage constraint
• cs.sv: value of the curtailment slack variable for each zone. This is the slack variable for the maximum curtailment constraint
• sl.csv: calculated storage level for all the generation units (included the units without any storage)
• dual*.txt: a set of text files with the dual values for all the constraints. The data is written to file directly from the model.dual Pyomo objects.