This presentation focuses on the work proposed to take advantage of technological advances of industry 4.0 – to change the approach of computational simulation and processing in real time – each of components involved in an industrial process. In this way, we’ll discover internal phenomenon which are complex to measure and are useful for predictive maintenance. This methodology is called Digital Twin in Plant Maintenance, and under several projects that had been made during the last years, it could verify an effectivity greater than 95% to predict and/or alerting abnormalities for assets from different industries.

The analysis of a building’s energy performance makes it possible to predict behaviors and mitigate associated risks. In order to carry out this analysis, it is necessary to obtain thermal models that are as close as possible to reality. In addition, modeling and simulation allow verifying its performance in a non-invasive manner, thus making it possible to propose energy optimization actions while enabling decision-making.

In this presentation, we’ll show the workflow of building and verifying pre-calibrated reduced order models, used for building optimal operation. First, the models will be generated automatically using TEASER (tool for fast generation of archetype buildings with low input requirements). Then, with Modelon Impact we’ll manipulate, simulate, and verify the pre-calibrated model. Finally, we’ll extract the model parameters to be used in the optimization environment GEKKO. We’ll conclude with the findings of the verification process.

The dawn of autonomy has arrived. While this topic ignites interest in the automotive industry, it has a fair share of difficulties. As humans, we are equipped with high complexity algorithms that allow us to control a vehicle safely. The task of replacing a human driver with an autonomous one is not straight forward. The autonomous driver needs to understand the behavior of the system under control to ensure a high-performance level. Hence, an accurate vehicle dynamics model is often required. The problem with such model is that accuracy comes at the expense of computational speed. In this presentation, a process is proposed for optimizing tractor-semitrailer vehicle models to ensure a high accuracy while maintaining a low computational cost.

Replacing base load providers on the energy market with decarbonized, renewable alternatives increases frequency dynamics and operational flexibility on the grid. Large transcritical CO2 heat pumps, as part of the ETES system by MAN Energy Solutions, engage slow dynamics for thermal energy storage and fast dynamics as they qualify primary frequency reserve for grid stabilization purposes.

We’ll present the dynamic model of a full CO2 transcritical heat pump cycle implemented with the Vapor Cycle Library, ThermoFluid Library, and own models. Integrating multistage compression, motor cooling streams and bypasses yields an in-depth representation of the full fluid path. Validation against testbed data of a commercial heat pump shows an accuracy within 10%. The dynamics during a valve variation along constant turbomachine speed and fast load balancing via speed variation – where power consumption changes over 70% from nominal within 30 s – will be demonstrated to match the testbed behavior.

Complex engineering systems operate in a wide variety of operating conditions. Estimating the performance of these systems using simulation requires the development of models capable of simulating variable operating conditions over long time scales. Because the time scales are long, it is necessary for these models to be robust, accurate, and minimal in complexity. To demonstrate an application of this method, a model capable of long-time scale simulation of the operation of a concentrated solar thermal (CST) plant designed to produce steam was developed in Modelon Impact. The simulated performance of this plant was compared to results from a similar model in Python and showed close agreement. This study demonstrates that long time scale simulations developed using Modelon Impact can be used to estimate the performance of complex engineering systems over long-time scales and a wide variety of operating conditions.

This presentation covers the performance evaluation of an air-rock bed concentrated solar thermal (CST) system requiring a robust and accurate rock bed Thermal Energy Storage (TES) model. We’ll show a dynamic CST system model with a reduced-order thermal energy storage model developed in Modelon Impact to balance model accuracy and computational cost. In addition, a high-fidelity MATLAB TES model used to calibrate the Modelica TES model. The behaviors of the high-fidelity and mock models after simulated for 10 charging/discharging cycles (240 hours of operation) in a plant cold startup operation will be compared by plotting the temperature profile within the TES along the length of the TES vessel. The error quantified with the normalized root mean square errors (NRMSE) of the fluid temperatures at the inlet and outlet of the TES at thousands of simulation time steps. This study will demonstrate that the reduced-order mock model could capture the physical behavior of the high-fidelity model, while improving the simulation speed.

For vehicles, thermal management developments are a key technique for reducing CO2 emissions and improving passenger comfort. As regulations continue to tighten, thermal management systems remain critical for modernizing. This presentation will focus on modeling and simulating automotive thermal management systems using Modelon Impact, integrated libraries, and additional libraries to optimize and analyze varying thermal management system models and components to generate improved results. We’ll cover enhancing and standardizing reports generated from results, varying parameter studies to directly see outcomes, and validate test results to understand the parallel outcome to reality.

The integrated libraries for this example in Modelon Impact include Modelon’s Air Conditioning Library, Liquid Cooling Library, Vapor Cycle Library, and Heat Exchanger Library. Additionally, Dymola models from other commercial libraries which are integrated into the workflow as FMUs. This session will give you an easy approach to rethink the way you model your systems and how Modelon Impact can support you and your customers.

In a massive decarbonization scenario, characterized by an increasing penetration of RES within energy systems, the integration of green hydrogen solutions plays a crucial role. RSOC (reversible Solid Oxide Cell) technology are usable for electricity generation (fuel cell, SOFC) and for hydrogen production (electrolyzer, SO-WEL) mode, allowing the cogeneration of thermal energy. However, rSOCs need a complex BoP for the thermal management and advanced controls, able to guarantee high efficiency even at partial load mode as well as easy start-up and shutdown procedures. In this work, a complete system model has been developed, focusing to the control development of the system. The developed model enables the evaluation of a detailed system analysis and operation. The model can be a useful design tool for the upscale of the rSOCs technology in different scenarios and applications, and to test and optimize the control strategy.

Providing clean operation of ships working to support the rapid expansion of offshore renewable energy generation in European waters is a major part of the maritime decarbonization efforts. Due to the near shore location and short sailing duration, vessels operating in support of the construction, operation, and maintenance of offshore wind farms are ideal candidates for the demonstration of zero emission ship operation.

This presentation will present the modelling of a vessels tank and power system in Modelon Impact to determine the feasibility of operating offshore service and support vessels with hydrogen fuels such as liquid organic hydrogen carriers. The model considers various operational profiles of such vessels replicating typical offshore endurances ranging from daily operations to month long offshore endurance for various types of work, in combination with an LOHC tank model and fuel cell sizing to determine the actual offshore endurance achievable.

Fortran is a procedural language developed in the 1950s and was especially suited for numeric computation. Our company has used Fortran since the 1970s to model and analyze physics-based equations to understand the thermodynamics of our systems. With a diminishing number of engineers coming into the workplace that know Fortran, a newer, more object-oriented modeling language was selected as a replacement that could also help us toward a standard to exchange models with customers and suppliers (i.e., FMI). Through this transition, we can add new features to our libraries, generate multiple abstraction layers to have a common structure between optimization, steady state and dynamics, add on a more user-friendly modeling interface, and make our compiler and solver more robust.

Hybrid power systems are of increasing interest in industry today. Systems that generate hydrogen through renewable energy sources and using that hydrogen as an energy storage source for use in dispatchable power generation are being evaluated around the word. Models have been developed using Modelon technology to evaluate the detailed operation and produce techno-economic studies of their viability with studies evaluating performance such as ROI. A detailed model using real environmental conditions as inputs to determine years’ worth of expected operational data have been developed and studied. The system model has been used in optimization studies to determine optimal system sizing in energy generation, storage, operational times to maximize system performance and efficiency. Modeling also includes detailed operational models such as ensuring gas blending is kept stable and to provide insight into important control schemes such as Trip Avoidance. The use of the python FMI is also included in this presentation.

This presentation focuses on the successful collaborate between ENGIE and Modelon to successfully build a digital twin for predictive maintenance on a solar power plant. We’ll cover how a physics-based model was used in Modelon Impact and integrated with real-time measurement data and maintenance records. We’ll show how doing this allowed our team to monitor main assets performance, detect anomalies, measure the impact of any event on performance, and feed Machine Learning algorithms to optimize asset management.