Interview with Samer Chaaraoui from Bonn-Rhein-Sieg University of Applied Sciences, Sankt Augustin, on the process of modeling photovoltaic systems. In cooperation with other members of the EnerSHelF project, he works on the mathematical representation of different effects and events that influence the operation of the system
When I searched for academic articles on solar photovoltaic modeling and simulation, I have seen a lot of equations and mathematical formulas. Can you explain in a more accessible way what modeling of photovoltaic systems encompasses?
By modeling photovoltaic (PV) systems, we try to find a mathematical representation of effects and events happening inside a PV system. With this representation, we can simulate processes, such as the conversion of solar radiation to electrical power, in order to estimate solar yields – for instance for economic and ecological business case analyses.
Since the full representation of the real world is not viable, we try to find mathematical representations which are simple enough to be calculated quickly and are complex enough to give an acceptable result. Therefore, it is especially important to validate modeling results with real world measurements, to estimate the performance of the model.
You will find many approaches and equations, trying to represent PV systems, which range from simple physical equations to more complex methods. Each of these approaches result from different demands towards the use case, usability, and accuracy of the model.
Within the EnerSHelF project, our use case is the dissemination of PV-based system solutions for the Ghanaian health sector. Since weather conditions are quite different to Germany, we must understand effects and events to find fitting mathematical models for PV systems. These models are used to calculate the economic and ecological benefit of such systems and to estimate their reliability. Furthermore, control algorithms are developed with a correct representation of the Ghanaian environment. The high reliability standards at health facilities make this task even more challenging, due to the low error tolerance towards the developed mathematical representations.
To meet these standards, we look at so called PV-hybrid-systems, which combine PV generators with battery systems and diesel generators as backup devices. Each of these components and their interactions need their own mathematical representations.
Multiple work packages (WPs) of the EnerSHelF project are involved in the modeling process as shown in the diagram below. Can you explain – in short – how everyone contributes to the overall results?
Every sub-WP of WP 3 contributes in different ways. For example, the work of WP 3.1 allows us to understand the electricity usage of Ghanaian health facilities. In WP 3.2 on the other hand, the team develops the weather forecast algorithm, which we use to lower uncertainties in terms of the high variability of solar energy. This information contributes to the control algorithm and allows an optimized dispatch of the PV-diesel-hybrid system.
With this, we can design and size PV-based energy systems via the design tool developed in WP 3.3. The design tool uses the mathematical model to simulate the lifetime of the energy system. With the simulation results, we can estimate the reliability and the economic and ecological feasibility. Furthermore, it allows WP 3.3 to adapt control algorithm designs to the high reliability demands mentioned before. This helps to lower the costs and polluting factors during the lifetime of the energy system.
As a next step, the team of WP 3.4 will upscale the results to a wider geographical range, to quantify economical, ecological, and social benefits of such PV-based energy systems to a national scale. By this, they will develop a market-introduction strategy for PV-hybrid-systems. Finally, the results of WP 3.2 and WP 3.3 will give input for WP 2 regarding the potentials of PV-diesel-hybrid systems.
How do you use the results in practice when designing the solar PV system for the field sites in Ghana?
Briefly speaking, we can use our findings to estimate the electricity load – the overall amount of electricity that is needed – of a Ghanaian health facility. More precisely, the developed models will offer an accurate representation for carrying out simulations to estimate the reliability as well as the economic and ecological feasibility.
Additionally, health facilities and their higher authorities can use our results to better estimate investments towards PV-based energy systems. The design tool will be a great contribution towards the dissemination of PV-based energy systems since an open access availability is planned.
What are specific challenges in the context of Ghana regarding site and environmental conditions?
As said before, Ghana does not have the same environmental conditions we experience in Germany. The weather is mostly damp and warm and – depending where you live in Ghana – you either experience a lot of cloud formations towards the coastal part, or less cloud formations but more aerosols in the northern part. Both affect the solar yield in different ways.
Moreover, you distinguish between the rain and dry season, rather than winter, spring, summer, and autumn. In the rain season, you experience higher variabilities and a lower solar yield than in the dry season. Within the dry season, you also experience the so called “Harmattan”, where winds blow dust from the Sahara Desert over West Africa towards the Gulf of Guinea. The Harmattan takes place between December and March and results into dust storms and hazy conditions, which can visually be mistaken as smog resulting from high pollutions.
At what stage of the modeling process are you at?
We have tested first load-forecasting algorithms with promising results, which range from simple statistical representations to complex algorithms based on neural networks. However, we still need to obtain more load measurements from other health facilities to validate these results properly. Furthermore, the design tool developed can now perform simple simulations and dispatch strategies. After properly developing the underlying models, we can start to develop the control algorithm for PV-diesel-hybrid systems. The weather forecasting model development is also on a good track. However, to validate the results, we need site measurements. Due to the Covid-19 Pandemic, the site measurements are delayed, which has a relatively low impact on our timeline but in long-term, we need those measurements.
When do you expect to be finished?
It depends on the current situation regarding the Covid-19 pandemic. The last step of our work is the validation of the models with real world measurements, which cannot be done right now. However, we are optimistic that we can optimize the development of the algorithms and tools, so that we can postpone the validation process and deliver the results on time.
Samer Chaaraoui is a Research Fellow at the International Centre for Sustainable Development (IZNE), Bonn-Rhine-Sieg University of Applied Sciences, Sankt Augustin. He holds a master’s degrees in Renewable Energy and a bachelor’s degree in Mechanical Engineering from the University of Applied Sciences Cologne, in which he worked in several laboratories and institutes. He also worked in an international corporation based in Germany prior to his work at the IZNE, prior to his return to the science domain. His main research interest includes control engineering, artificial intelligence and the promotion of alternative energy resources.