Project Directory Structure

The following figure gives a sample project directory structure. The following section go into more depth for the files stored in each section. See the Model Resources page for more information of the setup files that go into a project directory.

MiGRIDSProjects/ProjectName/InputData:

/Components:

All specific xml component descriptor files. For example, 'gen0Descriptor.xml'. These are created by running the function fillProjectData.py. This grabs user inputs from the /Setup/UserInput directory.

/Setup:

Has the project setup xml file. This is created by running the function fillProjectData.py. This grabs user inputs from the /Setup/UserInput directory.

/Setup/UserInput:

This contains all the csv files currently used by fillProjectData.py to populate the xml file used as inputs to run the simulations and by runDataImport to import data from the input data files. These will be replaced by the GUI at some point. The files include

• componentInformation.csv: a list of components and their power levels. This is used to fill the componentsNames tag of the projectSetup.xml file.
• generalTagInformation.csv: a list of tags, attributes and their values to fill into the projectSetup.xml file.
• timeSeriesInformation.csv: connects columns in input data files with components specified in the componentInformation file. (Maybe this could be ported over to componentInformation.csv)
• componentInformation.csv: These have tag, attribute and value information to fill into the xml component descriptor files (save in /Components). For example ees0TagInformation.csv.

/TimeSeriesData/RawData:

Contains the raw timeseries files.

/TimeSeriesData/ProcessedData:

Contains the processed time-series data files. These are created by running runDataImport.py. These are used to run the simulations.

MiGRIDSProjects/ProjectName/OutputData:

/SetX:

Contains the setup and output data for a set of simulations. X refers to the set number. Files include:

ProjectNameSetXAttributes.xml:

This file contains the changes to attributes in the xml component descriptor files in the project input (MiGRIDSProjects/ProjectName/InputData/Components) will be run for this set of simulations. Multiple values can be given for each attribute. Multiple values for one attribute must be separated by commas and no spaces. For example in the following code, multiple values are given for ees0.PInMaxPa.value and ees0.ratedDuration.value. The same value can be written to two attributes by writing the tag and attribute, separated by a period, to be copied in the value field. For example, in the following code, ees0.POutMaxPa.value will be identical to ees0.PInMaxPa.value. Simply copying the values 25,50,75 into the ees0.POutMaxPa.value field would result in 3*3=9 simulations for each possible combination of ees0.POutMaxPa.value and ees0.PInMaxPa.value. Attributes that require multiple values for one instance of a component (such as a fuel curve) cannot be changed here. A new component would need to be created and placed in MiGRIDSProjects//ProjectName/InputData/Components.

   `<compName value = "ees0 ees0 ees0 "/> 

    <compTag value = "PInMaxPa POutMaxPa ratedDuration"/> 

    <compAttr value = "value value value"/> 

    <compValue value = "25,50,75 PInMaxPa.value 3600,1800,600,300,7200"/>`

It also lists changes to the xml project setup file (located in MiGRIDSProjects/ProjectName/InputData/Setup). Multiple values are not allowed to vary between simulations in one set. Multiple values are allowed when it is required by the attribute, such as componentNames. Running the function generateRuns.py will generate a set of simulation runs using the information in this file.

setXComponentAttributes.db:

Running generateRuns.py will generate this SQL database with the table compAttributes. The columns in the table correspond to the component attributes that were changed from the base case values listed in the xml component descriptor files in MiGRIDSProjects/ProjectName/InputData/Components. Column headers list the component name, tags and attributes separated by ., for example ees0.PInMaxPa.value. The rows correspond to the run number. Each row has the values used for the attributes that are different from base case. The last two columns are called started and finished. A 1/0 in each column indicates the simulation run has/has not started and finished, respectively.

/SetX/RunY/Components:

This contains the xml descriptor files for the components used this simulation run. These files are identical to the base case xml descriptor files, located in MiGRIDSProjects/ProjectName/InputData/Components, except for the changes specified in /SetX/ProjectNameSetXAttributes.xml and summarized for each run in /SetX/setXComponentAttributes.db. Files are named componentNameSetXRunYDescriptor.xml. For example ees0Set2Run5Descriptor.xml

/SetX/RunY/OutputData:

This contains the net cdf time-series output files for this simulation run. They are named componentNameVariableNameSetXRunY.nc for example gen0PSet2Run5.nc.

/SetX/RunY/Figs:

This contains figures of plots of the time series for this run.

Project flow

This section goes over the flow of a project. See the tutorials page for projects that have already been set up for examples.

Import time series data

Run runDataImport. This processes raw time series data and saves the processed data into the InputData/TimeSeriesData/ProcessedData.

Create component descriptors and project setup files

Run fillProjectData. This creates xml component descriptor files and saves them in InputData/Components . These represent the base case for each component that is currently in the grid as well as components that will be added in the simulations.

Running fillProjectData also creates the xml project setup file which is saved in InputData/Setup. This gives the base case setup of the grid, including which components are run in the base case.

Grid Controller/Optimizer (not finished)

This should create a set subdirectory in OutputData and create a ProjectNameSetXAttributes.xml file listing which components to run and any changes to their attribute values from the base case. Alternatively, to manually run simulations, create a Set directory in the OutputData folder and create and fill a ProjectNameSetXAttributes.xml file.

Generate the simulations

Run generateRuns to generate all the run subdirectories in OutputData\SetX. It populates them with xml component descriptor files with the values for that run.

Run the simulations

Run runSimulation to run each simulation. Multiple instances of this can be run in parallel. time series net cdf files are saved in the run directory.

Get meta data (not finished)

Run getRunMetaData to get a performance summary for the simulation run. This saves the results into an SQL database in the set directory named setXResults.db. Inside the database, there is a table called Results which has the column headers: 'Generator Import kWh','Generator Charging kWh','Generator Switching','Generator Loading','Generator Fuel Consumption kg','Wind Power Import kWh','Wind Power Spill kWh','Wind Power Charging kWh','Energy Storage Discharge kWh','Energy Storage Charge kWh','Energy Storage SRC kWh'. As other components are added to the simulation, they will need to be added to this table. A possible change in the future is to make separate tables for each component. The row indices correspond to the simulation run number.

Back to Controller/Optimizer

The controller, optimizer or user can then decide which changes to make, create a new set directory with those changes, and repeat.