Concentrated Solar Power (CSP) plants capture abundantly available solar radiation and offer a renewable energy supply. The optimal design and operation of air coolers in such plants is key to their efficiency, however their dynamic performance is impaired due to fouling and variable external conditions. First, the air cooler section design is optimally designed in a modular way, enabling their use as a network of exchangers that can be operated independently. Then, exploiting the extra degrees of freedom, their simultaneous optimal operation and cleaning scheduling over 10 years is considered using a suitable formulation, with an economic objective function. Seasonal variability and operating conditions uncertainty are included. The resulting mixed-integer non-linear programming (MINLP) problem is solved efficiently using a Benders decomposition approach. Results are presented for two realistic CSP plants in Almería (Spain) and Dubai (United Arab Emirates), which have different external and weather conditions. The optimal solution reduces operating cost and defines optimal cleaning policies which are location dependent.

集中式太阳能发电站（CSP）可以捕获大量可用的太阳辐射，并提供可再生能源。在此类工厂中，空气冷却器的最佳设计和运行是其效率的关键，但是由于结垢和外部条件的变化，其动态性能会受到损害。首先，空气冷却器部分的设计以模块化方式进行了优化设计，使其可以用作可独立操作的交换器网络。然后，利用具有经济目标功能的合适配方，考虑利用额外的自由度，在10年内同时实现最佳操作和清洁计划。包括季节变化性和操作条件不确定性。使用Benders分解方法可以有效地解决由此产生的混合整数非线性规划（MINLP）问题。给出了阿尔梅里亚（西班牙）和迪拜（阿拉伯联合酋长国）的两个现实的CSP工厂的结果，这些工厂的外部和天气条件不同。最佳解决方案降低了运营成本，并定义了取决于位置的最佳清洁策略。