Success Story: Francis turbine draft tube modernization

NCC presenting the success story

Croatian Competence Centre (HR HPC CC) provides end users from scientific and higher education communities, various industries and public administrations, access to innovative solutions adapted to the level of maturity of national and European High-Performance Computing (HPC) infrastructure. HR HPC CC helps strengthen existing and develop new national competencies for High-Performance Computing (HPC), High-Performance Data Analytics (HPDA) and the area of Artificial Intelligence (AI).

Industrial Organisations Involved:

HEP is the national power company in Croatia. It manages several subsidiaries with different duties related to the production, delivery and distribution of electric energy. The scope of the company’s activities is vast. HEP Proizvodnja, a subsidiary, is tasked with the management and governance of the power generation facilities.

Technical/scientific Challenge:

Industrial partner sought a technical solution to improve the efficiency of a hydroelectric power plant with minimal investment / alterations. Although the obvious solution would be to modernize the runner of the turbine, it was questionable whether the long-term gains would justify the expenses. Solution was seen in the modernization of the draft tube which can be done quickly, during the scheduled maintenance of the plant. Based on provided data, it was necessary to develop a computational model which was used the conduct complex CFD simulations in order to determine an optimal solution.

Solution:

A numerical model was created and verified to ensure the validity of the approach. Based on calculated data for the entire turbine, a segmentation approach was adopted and only the draft tube was considered. Geometry parametrization and optimization were considered in phase I, in order to generate an initial, optimized design. In phase II, machine learning was employed to create a meta-model that can subsequently be easily employed with optimization algorithms or other approaches/goal functions to determine the best design based on specified criteria/constraints. Due to the complexity of the numerical modelling in general, as well as the scope of the optimization problem, the use of HPC resources was mandatory, as was the tuning and development of the noted meta-model.

Business impact:

The proposed methodology can serve as a proof of concept for other sections of the power plant or can be used to design new plants. This means that additional improvements can be made. Due to the complexity of the problem, conventional systems cannot be employed to calculate the problem and provide an optimized design. However, by utilizing HPC resources, simulations and the optimization process were completed in an acceptable time frame. Furthermore, the use of ML and the development of the draft tube meta-model significantly reduces computational time and effectively eliminates the need for the HPC in subsequent analyses/optimization processes. Obtained results led to the improvement in the efficiency of the turbine and as such are being considered as potential solutions. Based on the presented methodology, a more in-depth assessment and efficiency improvement steps are being conducted.

Benefits:

  • Approach can be employed for other hydroelectric power plants
  • Approach is flexible in terms of constraints/requirements and is applicable to other turbine segments i.e. this proof of concept methodology can be subsequently used on other sections of the turbine
  • Increased turbine efficiency (production of electricity)

SUCCESS STORY # HIGHLIGHTS:

  • Keywords: Computational Fluid Dynamics, High Performance Computing, Draft Tube, Machine Learning, Optimization
  • Industry sector: Mechanical engineering
  • Technology: HPC, ML

Suggested optimized designs

Contact:

Lado Kranjčević lado.kranjcevic@riteh.hr

This project has received funding from the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 951732. The JU receives support from the European Union’s Horizon 2020 research and innovation program and Germany, Bulgaria, Austria, Croatia, Cyprus, the Czech Republic, Denmark, Estonia, Finland, Greece, Hungary, Ireland, Italy, Lithuania, Latvia, Poland, Portugal, Romania, Slovenia, Spain, Sweden, the United Kingdom, France, the Netherlands, Belgium, Luxembourg, Slovakia, Norway, Switzerland, Turkey, Republic of North Macedonia, Iceland, Montenegro