Integrated planning and optimisation of municipal electricity and heating networks
|Software type||Planning, Balancing/Optimising|
|Term of project||Oktober 2014 - März 2018|
Between October 2014 and March 2018, the project consortium, consisting of the Rechenzentrum für Versorgungsnetze Wehr GmbH (RZVN) engineering office, the University of Constance and Constance University of Applied Sciences (HTWG), is focussing on developing planning and optimisation methods for designing hybrid municipal energy generation and distribution structures.
What’s it about?
The current optimisation of the energy networks is primarily segment-oriented and largely relates to the existing distribution of electrical and thermal loads. So far it has not been possible to depict the effects of decentralising the electricity and heat generation nor the effects that changing building technologies are having on the distribution networks. The optimisation methods for energy distribution systems previously known from practice and specialist literature usually relate to isolated and centrally fed electricity distribution grids or expansion plans for gas or district heating networks. The joint project, which is entitled "Hybrid planning process for supplying energy-efficient heat and electricity to urban distribution networks" (HYPV), therefore intends to overcome these system boundaries and model the stronger coupling of energy subsystems in future (electricity, heat, natural gas) and thus optimise the entire system.
What has been achieved so far?
So far RZVN GmbH and Constance University of Applied Sciences (HTWG) have developed preliminary network and building energy cost models for linear optimisation programmes and fed them with parameters and cost functions. They have also created a preliminary framework for a simulation and optimisation model that can be expanded as required. This enables the influence of external energy procurements to be taken into account, for which the first costing and technology scenarios have already been researched. In addition, RZVN GmbH has developed a preliminary tool for simulating necessary municipal electricity and heating network expansions through increased use of CHP plants and heat pumps. Parallel to this, the University of Constance is developing algorithms to optimise the building-related technology mix taking into account the necessary network expansion.
Rechenzentrum für Versorgungsnetze Wehr GmbH (Dr D. König, Dr P. Hensel), which is based in Dusseldorf and Constance, has more than 50 years of expertise in calculating, planning and sizing gas, electricity and district heating systems for approximately 400 municipal utility companies. It is in charge of developing the interdisciplinary optimisation model for determining the optimal network structure of the electricity grid as well as the supplementary gas and district heating networks (optimisation level 2). Here it can build on existing modelling approaches concerned with planning the expansion of gas and district heating networks in regards to specific buildings.
The following work packages are planned:
- Development of interfaces for acquiring network data
- Mathematical depiction of parallel network structures for electricity, gas, local/district heating for the optimisation tasks
- Solving and testing different optimisation methods for the networks belonging to the municipal utility companies involved.
Constance University of Applied Sciences (Energy Economics Group; Professor Göllinger) has experience in designing and implementing inter- and trans-disciplinary research projects. Its range of expertise covers the modelling of energy systems, systematic technology and environmental analyses as well as the decision-relevant implementation of scenario analyses and the processing of the results. Specifically, Constance University of Applied Sciences is designing all the basic models that supply the optimisation models used by RZVN GmbH and University of Constance with data; for example, models for standard building classes, taking into account that the building owners leverage the respective efficiency potential to varying degrees. In addition, it is defining suitable scenarios for developing technology costs and policy frameworks, and will subsequently collate the optimisation results in order to identify stable optimality areas.
Based on RZVN’s linear MILP model (Mixed Integer Linear Programming), the University of Constance (Professor S. Volkwein) is developing a nonlinear combined model that enables interdisciplinary optimisation of the networks. The efficient solving of MINLP models (Mixed Integer Nonlinear Programming) is an active research area in mathematics. In particular, the research group specialises in developing numerical algorithms for solving high-dimensional, non-linear minimisation problems, whereby it focuses on the resolvability of high-dimensional model instances such as occur in urban network structures.
Next steps, Milestones, Outlook
End of 2015:
- Determination of the relevant technologies
- Assessment of the market potential of future technologies
- Determination of the data structures and grid models
- Documentation of the current state of the algorithms
- Literature review
End of 2016:
- Initial conclusions on the economic viability of the respective technologies for standard building types
- Testing the MILP model to determine the network structures and the local technology distribution
- Linking the results of the MILP model with hydraulic network calculations in order to validate/adapt the results
- Selection of suitable MINLP solutions
End of 2017:
- Completion of the modelling and development work
- Detailed application of the system to the model cities of Constance, Sindelfingen and Dusseldorf
- Validation of the system proposals with practice partners
- Depiction of the transferability of the results to other cities
A comprehensive simulation and optimisation tool using scenario-based, cost-minimised, hybrid target network calculations shall enable municipal utility companies and other decision-makers to make decisions on an informed basis and with reduced risks. To achieve this, the results of the cooperation partners shall be linked via a joint interface.