Integrated AC & DC Design & Analysis
Model-Driven SCADA, EMS, PMS, ADMS & SAS
A Unified Digital Twin PlatformDesign, Operation, and Automation
Typically, the procedure for predicting the energy yield of a PV plant using time-step (hourly or sub-hourly) simulation software will consist including estimated losses, site conditions & historical irradiance. These are good for estimation but fail to include detailed AC&DC losses, auxiliary power, grid availability and disruption, grid compliance loss, power plant controller performance, etc. Developers using non-model driven tools are usually not aware of or do not consider several additional losses and do not calculate a long-term yield prediction over the life of the project with uncertainty analysis. Both are essential for potential project financiers. An investor will usually look for a higher level of confidence in the energy yield prediction, typically expressed as the P90 value, or the annual energy yield prediction that will be exceeded with 90 percent probability.
The ultimate aim of the designer is to design a plant that maximizes financial returns by minimizing the Levelized cost of electricity (LCOE). LCOE is the present value of the price of the produced electrical energy (usually expressed in units of cents per kilowatt-hour), considering the economic life of the plant and the costs incurred in the construction, operation, and maintenance, and the fuel costs. LCOE does not take into account the site environmental factors other than those considered in the historical data such as panel soiling that may be accelerated due to unforeseen conditions, technical faults such as a faulty inverter, or damaged PV panel due to dust or rocks and changing site conditions such as vegetation. A power plant controller must be able to detect and adapt to these changing conditions.
Independent power producers can commercially operate if they are satisfying all the requirements stipulated in the national grid connection code. Consequently advanced plant controllers must be implemented not just in the operations phase but also in the project design phase. The typical control requirements are in terms of megawatts and mega-VARs, (active and reactive power). Optimally, a solar PV plant appears to the grid as a single, unified source of power. The goal is to maximize power output (and, therefore, revenue) while supporting a stable and reliable grid using a configurable automated controller.
The logic used for the power plant controller must be capable of precision tuning and control again during the design and operations phase. It is possible that assumptions during design stage lead to a slow or fast responding controller that may need to adjusted during commissioning phase. Having access to the model, the controller logic allows for updating simulation models and increasing the confidence level that similar performance will be achieved in the field.
The logic used for the plant controllers during design stage must be the same as that used for operation phase for another reason. You run the risk of meeting grid code requirements during design stage but not during operations if you utilize a controller logic in simulation mode, that is not commercially available to be used in operations. The cost to design, rewrite and maintain two controller logics can delay commissioning and increase start-up costs.
A power plant owner/operator can easily lose their license to operate or incur financial losses should their plant fail to meet ongoing/evolving operational grid code requirements. Therefore there is a need to continuously monitor the power exchanges, power quality, and grid conditions and proactively adjust plant performance should it not be within grid code requirements.
A single point of management is needed for remote monitoring of plant performance. A monitoring system is an essential part of a PV plant. Monitoring devices are crucial for the calculation of liquidated damages (LDs) and confirmation that the EPC contractor has fulfilled its obligations. Automatic data acquisition and monitoring technology are also essential during the operational phase to maintain a high level of performance, reduce downtime, and ensure rapid fault detection. A monitoring system allows the yield of the plant to be monitored and compared with theoretical calculations and raise warnings daily if there is a performance shortfall. Faults can, therefore, be detected and rectified before they have an appreciable effect on production. Without a reliable monitoring system, it can take many months for a poorly performing plant to be identified. This can lead to unnecessary revenue loss.