(434a) The Use of Planning in a DRTO Methodology for a Solar Thermal Plant | AIChE

(434a) The Use of Planning in a DRTO Methodology for a Solar Thermal Plant

Authors 

Sochard, S., Université de Pau et des Pays de l'Adour (UPPA)
Marias, F., Université de Pau et des pays de l'Adour
Reneaume, J. M., École Nationale Supérieure en Génie des Technologies Industrielles
Carrillo le Roux, G., University of São Paulo
Serra, S., Universite de Pau et des Pays de l'Adour, E2S UPPA, LaTEP
A solar thermal plant can supply heat to industries and district heating networks without direct CO2 emissions. In such a plant, both the energy source and the heat demand are variable, making its operation challenging. Thermal energy storage can help to decouple the solar heat production from the consumption, giving more flexibility to the solar thermal plant by further complexifying its operation. Finally, weather and load forecasts are not always accurate, so the optima operation cannot be fully planned in advance, since it might not be optimal in the actual environmental conditions. Nowadays, most solar thermal plants are controlled using logic control rules and basic controllers (Camacho et al., 2007). Dynamic optimization has been applied to solar thermal plants, such as in (Scolan et al., 2020), improving the economic performances of the heat production. However, offline dynamic optimization cannot adapt the optimal strategy in real-time.

The methodology developed in this work is a Dynamic Real-Time Optimization (DRTO) in association with a planning phase o improve storage management, as suggested in (Untrau et al., 2022). The planning level is an economic dynamic optimization which determines the optimal storage management policy a few days ahead based on weather forecasts, benefiting from a longer-term strategic vision. The DRTO level adapts the optimal operation of the solar thermal plant, in terms of optimal flow rates in the different parts of the plant (solar field, charge, discharge, supply, etc.), to the current disturbances and updated forecasts. The DRTO can include a term on storage management in its economic objective function. Since it is an online methodology, it was tested on a detailed simulation model undergoing the actual weather conditions and providing feedback to the DRTO level. In this work, the control is assumed perfect, and no controllers are modeled.

The methodology has been tested in several scenarios to determine the best way to integrate storage management in the DRTO objective function. The scenarios are different test periods chosen in different seasons. Real values for the weather forecasts and measurements are used for the year 2021 in the city of Trappes, France.

First, the DRTO level helps to correct any error in the weather forecasts and adapts the optimal operation to the actual conditions. Moreover, it helps to reduce the model error propagation due to simplifying assumptions in the optimization model. Thus, DRTO generally outperforms offline dynamic optimization.

An important aspect of this work is to assess the best way to integrate storage management in the DRTO objective function. If no objective on storage is included in the economic objective function minimizing the operating costs, the solar thermal plant performances are low because no excess energy is stored in anticipation of a future period without enough solar irradiation. Thus, the total energy supplied to the heat consumer is low. It was found that maximizing the stored energy at the end of the DRTO time horizon is a good choice in most cases because it maximizes solar energy harvesting and extends the duration of solar heat supply. However, when there is a risk of overheating, which happens when the storage tank is full and there is more solar irradiation than needed to meet the heat demand, it is preferable to follow the planned storage management policy. Thanks to a longer term strategic vision, planning is more likely to avoid overheating situations. Thus, minimizing the difference between the planned stored energy and the actual stored energy at the end of the time horizon of each DRTO helps to prevent overheating, if the weather was correctly predicted.

In future work, more test periods should be evaluated to determine a criterion allowing the methodology to switch between the DRTO maximizing the stored energy and the DRTO following the planned energy when needed.

References:
Camacho, E. F., Rubio, F. R., Berenguel, M., & Valenzuela, L. (2007). A survey on control schemes for distributed solar collector fields. Part I : Modeling and basic control approaches. Solar Energy, 81(10), 1240‑1251. https://doi.org/10.1016/j.solener.2007.01.002

Scolan, S., Serra, S., Sochard, S., Delmas, P., & Reneaume, J.-M. (2020). Dynamic ptimization of the operation of a solar thermal plant. Solar Energy, 198, 643‑657. https://doi.org/10.1016/j.solener.2020.01.076

Untrau, A., Sochard, S., Marias, F., Reneaume, J.-M., Le Roux, G. A. C., & Serra, S. (2022). Analysis and future perspectives for the application of Dynamic Real-Time Optimization to solar thermal plants : A review. Solar Energy, 241, 275‑291. https://doi.org/10.1016/j.solener.2022.05.058