Questions & Answers (Q&A)

How can I install it?

What PC do I need?

  • openTEPES requirements depend on the case study

  • The main dimensions to consider are:

    Time (periods, load levels). For example, select one year out of every five, use representative days, or define a 2-3 hour time step instead of 1 hour

    Generating units. Use generation types instead of generating units

    Network (nodes, lines, transport or DC power flow, ohmic losses). Use a reduced network (corridors) with fewer nodes and lines, transport model instead of DC power flow, or ignore ohmic losses

    Stochasticity (scenarios)

    Binary investment decisions (candidate generators, storage, transmission lines), operation decisions (commitment, startup, shutdown), and mutual exclusivity between units

    Sectors to consider (electricity, heat, hydrogen, hydro basins, etc.)

  • As a rule of thumb, a linear optimization problem requires 1 GB of memory for every 1 million rows. For a mixed-integer linear optimization problem, the requirements are much higher, and they depend on the number of binary variables and the number of constraints that include them.

  • So, depending on the size of the optimization problem and the available memory, you may or may not be able to run it on your PC. As an example, the case studies provided can be run on a laptop with 32 GB of memory.

First steps

  • openTEPES is provided with some case studies. Each one has varied characteristics

  • Check that you can run them on your PC. Some of them may not run if you have a PC with little RAM

Potential issues

  • Check that any item you define (generator, node, technology, area, etc.) is included in the corresponding dictionary (oT_Dict_Generation, oT_Dict_Node, oT_Dict_Technology, oT_Dict_Area, etc.). Otherwise, the optimization problem will not run

  • openTEPES is implicitly a network model. Therefore, at least two different nodes and one electrical transmission line connecting them must be defined

  • At least one generator must be defined. A thermal generator has a variable cost greater than zero (the product of fuel cost times the linear term ≠ 0). A variable renewable energy unit has zero variable cost (the product of fuel cost times the linear term = 0)

  • Check the correspondence between the resources you define for the system in the case study to be run and the nodes where they are meant to be located. The location within the grid of any resource defined must be specified.

  • Check that all the demand in the system within the case study is in a node that can be served with the energy output of some resources located in any node that is connected/to be connected to the former, possibly including generation representing energy not served (ENS).

Some tips

On building an appropriate set of input data files for a case study that can be run on openTEPES:

  • How can I significantly reduce the size of the case study for testing purposes?

    Delete the duration (column D in oT_Data_Duration) of the load levels you want to ignore. For example, if you delete the duration beyond the first 168 hours, you are considering just the first week of the year in the case study

    Set a time step of 2 or 3 hours in the oT_Data_Parameter file. This will reduce the number of load levels by a factor of 2 or 3

    For working with representative stages, see, for example, the 9n7y case study

  • Relaxing the binary condition of binary variables

    In the oT_Data_Option file, you can force or relax the binary condition of the binary variables.

    They can also be relaxed individually for each generator or transmission line in oT_Data_Generation and oT_Data_Network.

  • How to ignore a scenario or a period

    Assign probability 0.0 to the scenario in oT_Data_Scenario or weight 0.0 to the period in oT_Data_Period

  • How to ignore a generator

    Do not assign a node to it in oT_Data_Generation, i.e., leave the node column empty

    Set the initial period where the generator can be in operation beyond the year of study in oT_Data_Generation

  • How to ignore a line

    Set the initial period where the line can be in operation beyond the year of study in oT_Data_Network

  • All the empty cells of the CSV files are replaced internally by openTEPES with 0.0

    Therefore, if you want to set the generation of a solar PV to 0.0 at night, then you must enter a small value, 0.000001, that will be replaced internally by openTEPES with 0.0. If the cell is left empty, openTEPES will consider the rated capacity of the solar PV unit defined in oT_Data_Generation as the generation at night.

  • There is no need to include the column for a certain resource (generator) within an input data file if there is no data to be defined for this resource within that file. Empty columns do not need to be included in some input data CSV files, including oT_Data_VariableMaxGeneration, oT_Data_VariableMinGeneration, oT_Data_VariableMaxConsumption, oT_Data_VariableMinConsumption, etc., but at least an empty column with the name of the resource must be included in the file header

On analyzing output results:

  • Make sure the problem-solving process has been successfully completed, reaching optimality (the console log and the solver log file provide information on this).

  • If the problem-solving process has not produced an optimal solution, check if the system conditions defined within the input data files are too tight, i.e., the system may not have been provided with large enough flexibility for the model to find the optimal problem solution. If this is the case, some problem constraints could/should be relaxed to allow the model to compute an optimal solution.

  • Check the level of the overall system variables in the output energy balance files (e.g., oT_Result_BalanceEnergyPerArea, oT_Result_BalanceEnergyPerTech) to assess whether they seem to make sense. Focus first on certain specific variables, including the ones that follow:

    Non-served energy amount

    Amount of spilled and curtailed energy

    Overall output by technology, if you have some reference levels for this to compare to

  • Whenever the level of some variables at the system level does not seem to be reasonable, check the output data file for the energy balance at the area (country) level, to try to identify in which area within the system the problem may be located

On analyzing many scenarios:

  • If there are several scenarios to be analyzed, and they are not linked by investment decisions, the model automatically solves one scenario after the other and presents the results for all the scenarios in the output data files. In this case, the results for each scenario can be easily identified by the scenario name in the corresponding column header of the output data files. This can be limiting if there are many scenarios and the system is large.

  • In that case, it may be more convenient to run the scenarios one by one (each scenario located in a different folder), and save the results for each scenario in the corresponding folder. This can be done by running openTEPES simultaneously in several consoles.