Nobody can stop the wind, cold the sun or plug a geyser. This means that a power plant based on any type of renewable resource, whether wind, sun, or geothermal heat, is not controllable straight at the source. Any power plant that operates any renewable and uncontrollable resource can potentially feed the grid with an unpredictable amount of energy at any time.
Production of energy from renewable sources is expected to nearly double in the next 20 years all over the world. It’s great news for the environment but also a crucial challenge for the industry. Today’s 20% share of energy produced from uncontrollable resources is a thorny problem for the stability of the grid and it will exponentially sharpen in 20 years.
Therefore, control is required and can only come via tailor-made, and especially intelligent, software. A software that simulates the operation of multiple plants and that allows to forecast their future behavior, adjusting their production according to needs. In fact, more often than not, energy supply and demand do not match.
An Energy Management System (EMS) is a system capable to transform an uncontrolled output into a virtual power plant which produces a controlled output. To reach this objective, the EMS must take into account a number of factors such as current and forecasted weather conditions, power plant failures, downtimes, planned maintenance, market price and demand forecast, and ultimately makes decisions in a matter of seconds.
An ideal EMS will therefore come with an extensible rule engine that allows to define custom alarms based on mathematical and logical expressions that combine together thresholds, real-time conditions (internal temperature, rotations speed, hydraulic pressure, vibration) and historical data (production, wind speed, solar irradiation).
EMS decision-making process is based on A.I. and an artificial model of reality: a digital twin of the system, a simulator that allows to replicate reality. It is purposefully made to be interacted with, the operator can experiment and simulate any condition on the digital twin and see the results applied to reality.
The image below illustrates an EMS which regulates the power production of multiple wind, solar and hydro power plants to produce a fixed output and meet market demand.
The EMS is considered to be the “central brain”. It receives real-time data from “edge brains” which collect data and send inputs to the EMS regarding a number of factors such as how much power plants are producing, how much wind there is or about stopped turbines.
EDGE systems talk to plants through first-level SCADA systems, or standard data interfaces such as IEC 61850 -104.
In addition, to improve its predictions and take real time decisions, the EMS collects and elaborates other kinds of inputs which allow for:
- Planned and unplanned stops. Power plants cannot be assumed to work regularly and constantly, due to errors which cause them to shut down such as generators overheating or hydraulic systems malfunctions. EMS collects data coming from these alarms and statistically predicts future occurrences. Power plants can also stop according to a regular maintenance plan which can be provided as an input to the EMS.
- Weather and power forecast. Predicting future production level is fundamental to take timely decisions, to start and stop a power plant, or curtail the power output to a certain level. Because the power production depends from the weather conditions an EMS must have the capability to forecast weather conditions and the effect that those conditions will have on power production levels.
- Demand level and market price. Predictive analysis, based on machine learning algorithms, forecasts energy market demand and predicts load and prices. This with the purpose to manage power flows and minimize electricity generation costs.
Those are just some of the element that a modern EMS must take into account in real-time to control and orchestrate the power production, which would otherwise be unpredictable.
In conclusion, implementing an Energy Management System will optimize assets, anticipate and respond to system disturbances and will allow power plants to operate resiliently against attack and natural disasters, providing power quality.