FREM Article Series: Geodata in Energy System Modeling

This article starts the 6-part series on the FfE Regionalised Energy System Model FREM. In the coming weeks, the series will be expanded to include the core elements of FREM, such as various regionalisation models and data.

Overview of the topics in the series of articles concerning FREM
  1. Development of an Energy-related Geo-database
  2. Regionalizations for the Network Development Plan
  3. The Wind Scenario Tool WiSTl
  4. Modeling the Expansion of Ground-Mounted & Roof-Mounted Photovoltaic Systems
  5. Weather Data
  6. Open Data
Figure 1: The development of FREM from the first data sets to the comprehensive geo-database today.

From energy system model to comprehensive database

FREM, which was launched in 2007 as a building dataset for modelling heat and electricity consumption, is constantly being developed and expanded by FfE. Since the beginning of development, the database has grown continuously through the constant integration of new data and the focus is constantly being expanded to include new applications (see Figure 1). Today, FREM as a comprehensive geodatabase forms the core of the FfE’s data landscape. The rapid development in the availability and processing possibilities of spatial data in recent years has made geographical information systems (GIS) and geodata increasingly relevant for energy system modelling as well.

Through the development of interfaces to the other FfE models such as ISAaR and GridSim, FREM has become increasingly interconnected in the FfE model landscape and, as a data management system, is now a core component in almost all FfE projects. Since 2016, data exchange between FREM and external databases has also gained in importance. The increased provision of baseline data will also be one of FREM’s goals in the near future. In this context, the OpenEnergy Platform (OEP), among others, is of central importance, as open data on the energy system is playing an increasingly important role. For this reason, the FfE actively supports the OEP through FREM and is also constantly expanding its own data portal, opendata.ffe.de.

Today, FREM is a comprehensive geodatabase built on the open source systems PostgreSQL and PostGIS, with spatial and statistical data on the energy system (Figure 2). These data originate mainly from open data sources. Primary and derived data are stored in thematic schemas such as renewable energy, weather models, time series, power plants, statistics and geographic data. The consistent structure of the database is key for fast modelling and scenario calculation to solve various energy issues. Through the geo extension PostGIS, FREM can be used like a geographic information system (GIS) that puts energy data into a spatial context. It is possible to derive data in different regional resolutions with SQL queries and export the results in different data formats for cartographic and statistical visualisations.

Figure 2: Overview of the Strucutre of FREM

The IT structure of FREM

The heart of FREM’s IT structure is the powerful Mercator server, which hosts the geo-database and runs under Ubuntu Linux. Written SQL code is versioned in GitLab and sent directly or in parallel via SQL parser to Mercator, which can thus make full use of its 24 cores. Careful physical separation of the operating system, geo-database, data and logs on different raids further reduces computing times. A comprehensive backup strategy with partial and full dumps protects against devastating data loss in the event of damage. In addition to its function as a weather data store, Herodot, Mercator’s predecessor, acts as a time-delayed mirror of the FREM database to enable short-term data recovery without costly backup and restore operations. The connection to another geo-database hosted on an external server represents the FREM interface “to the outside” and supplies the Open Data Platform with data that can be freely obtained from there via a REST API.

Use cases:

The combination of the PostGIS geo-database with comprehensive energy industry data thus results in a flexible energy system model with high temporal and spatial resolution. This model is used as a rich data pool in most of the FfE’s projects, e.g. to survey renewable energy potentials, power grid analyses for different voltage levels, and future electric vehicle infrastructures. Some of this core work of FREM will be presented in the upcoming contributions of the series.


Further information: