Simulated load profiles of a laboratory building for air-conditioning cooling, room and process heat.
© RWTH Aachen

Depiction of the superstructure of an energy supply system. The blue network represents the hot/cold water system, the red network represents the electricity supply.
© RWTH Aachen

Schematic illustration of a supply network in the TOP-Energy programme.
© Gesellschaft zur Förderung angewandter Informatik e. V.

Supply lines on Knapsack chemical park
© InfraServ GmbH & Co. Knapsack KG
2 / 3

Optimisation with sOptimo

Pre-optimisation is implemented in a simplified method of calculation with constant efficiency factors and variable efficiency factors are used for subsequent re-adjustment. For a shorter computation time, the researchers use generator models with constant efficiency factors. This simplification drastically reduces the solution duration. The option to use variable partial load efficiency factors for greater accuracy results in a higher computing time for optimisation, in particular for complex systems. This can be counteracted through optimisation using an “optimality gap”, which specifies how far the final target function value may deviate from the theoretically possible, optimal target function value in per cent. A time limit can also be set that defines the latest possible point at which optimisation can be aborted. The two alternatives usually produce a capital value that is close to the optimum. The generator structure barely changes when the result of the calculation using an “optimality gap” is compared against the mathematically optimal solution. Where the solution duration is comparatively short, nearoptimal solutions are consequently delivered. Several different supply structures with almost identical capital values emerged in testing.

The new tool has been tested on a real life example with the support of the developers. The testing revealed the tool to be capable of mapping real life structures too in their full complexity and evidenced the capability of the optimisation processes to identify improvement potential. Several additional practical issues, however, might be better handled if the programme were to include additional functions. One of the objectives in this regard is to be able to identify several alternative proposals with the best possible solution properties. A further challenge lies in factoring in several conflicting target functions, e.g. costs and primary energy use.

Simulation improves planning

To test the programme, the experts drew up an industrial site plan for a globally operating pharmaceuticals company and determined the optimal supply structure based on capital value. For anonymity, the experts transferred the industrial park’s buildings and roadways to the site plan of RWTH Aachen. The experts were able to utilise existing data on maximum cooling loads in the calculations. The researchers calculated the maximum heat loads and the load profiles for the required air-conditioning cooling and room heating with a thermal building simulation. The estimated process heating and power requirements were illustrated as load profiles. Load profiles were generated for 21 buildings (production, research, office, warehouse, canteen and data centre). The real life energy supply system with electricity, heating and cooling consumers and suppliers – boilers, CHP, absorption chillers and vapour-compression chillers – is mapped by a superstructure in the Top-Energy programme. All possible variants for supply of the energy required are coded in the superstructure. A subset is then selected from these variants in the optimisation. Compared to the real life system structure, the optimal system with two integrated CHPs saves 40 % of the costs of procurement and investment over a ten-year period.
The researchers achieved this by optimising the supply structure: Compression and absorption chillers, CHPs and boilers, connected to the public power and gas network. How long the computation takes depending on the complexity of the superstructure and the number of examined time steps was also looked at. Based on two generators with variable efficiency factors, it was possible to calculate models within a minute.

Translation into operational practice

InfraServ GmbH & Co. Knapsack KG is involved in the project in a supporting role as a practice partner in the industrial parks sector. Two industrial parks operated by the company near Cologne are being used as application examples. For these industrial parks, the software automatically drafts variants of the existing power system and analyses these. In the process, both the structure and the topography of the power system are optimised. Criteria used in particular are the costs and air-conditioning efficiency of the system. The optimisation software not only compares variants already considered by planners, it also develops and analyses entirely new ones.
The objective of ongoing efforts is to make the methodologically verified model utilisable for a wider circle of users. The aim is to extend the tool’s use to planners without proficiency in mathematical optimisation. This necessitates additional requirements with respect to the implementation of the method in TOP-Energy. The project has been funded by the German Federal Ministry for Economic Affairs and Energy. The developers hope to launch a follow-on project to develop the software further. It is hoped future functionality will allow complex real life systems to be depicted with monthly metering time periods or even daily averages. In addition to integrating multi-criteria optimisation, the hope is that the programme will be able to address problems defined by industry partners too.

Waste heat from refineries destined for district heating

Karlsruhe provides an intriguing example of how waste heat potential can be utilised by networking generators and consumers and large quantities of energy can be saved. The Oberrhein mineral oil refinery (MiRO) is the largest refinery in Germany. The facility is unable to make further economic use of waste process heat registering below 130 °C. This temperature, however, is sufficient for district heating supply. Since autumn 2010, Karlsruhe municipal utilities have been using waste heat sourced from a section of the MiRO plant for district heating supply. The waste process heat is taken from the refinery’s plants via compact plate heat exchangers and fed into the district heating network. The use of waste heat has increased the energy efficiency of the refinery by five per cent. The municipal utilities sourced around 300,000 megawatt-hours in 2013. This equates to the heat demand of approx. 25,000 households and saves 65,000 tonnes of CO2 annually.
Waste heat to be sourced from a second section of the plant from October 2015 is expected to supply a further 220,000 megawatt-hours. The German Federal Ministry for the Environment funded this pilot project as part of the climate protection initiative, spending five million euros. It was distinguished with the Energy Award in 2013.


BINE subscription

Subscribe to publication


Project management
RWTH Aachen, LTT

GFaI e. V.

Operator, energy service provider
InfraServ GmbH & Co. Knapsack KG

Operator, energy service provider
Siemens AG


Structural optimization of distributed energy supply systems
Link to the website of the research project of RWTH Aachen university

Learning Energy Efficiency Networks
Website of the LEEN GmbH