Key points
Power plants are heavy consumers of energy. To remain competitive as global deadlines for carbon-free emissions draw near, energy companies that generate power must find ways to become more efficient with their own energy use. For many, this means investing in new technology to explore the operational data of processes used in power generation.
Events during the past few years have strengthened a need to digitalise operations at power plants, according to research from McKinsey & Co. They include the COVID-19 pandemic, fluctuating fuel prices, a focus on renewable energy, and various regulations aimed at achieving Net Zero operations. Recent supply shortages and inflation also have accelerated the urgency for better energy management through digitalisation.
One of the ways the energy industry can benefit from these new technologies is the ability to leverage clues hidden in operational data. This sensor-generated data offers insights on energy consumption and identifies areas for possible improvement.
Engineers can use this information to find opportunities to increase energy efficiency. They also can use it to develop a more comprehensive energy management plan and to meet reporting requirements that demonstrate compliance with environmental regulations.
Operational Data and Energy Management
There are several ways that power plants can use operational data to improve energy efficiency. They include:
- Monitoring energy consumption: Using good periods of operation as a golden fingerprint, a monitor and alert system can be developed that looks for deviations from ideal parameters and notifies personnel with enough time to make corrections.
- Performing preventive maintenance: Fouling of critical systems can occur over time. When maintenance is performed too early, it leads to unnecessary downtime. But maintenance performed too late can lead to inefficient operations. In the worst situations, it can cause a plant shutdown and damage equipment. Operational data can help determine the right time to perform maintenance.
- Optimising asset performance: When process performance is degraded, it can lead to capacity losses and a variety of efficiency concerns. In a power plant, process anomalies also can cause unwanted substances (such as sulphur emissions) to be released into the environment. Information on pipe network anomalies, poor quality raw materials, and a variety of other factors that prevent Overall Equipment Effectiveness (OEE) can be found in operational data.
Traditionally, analysing operational data was a job for a data scientist or central data team. This caused a delay in implementing efficiency improvements because data scientists are overbooked with improvement projects. Statisticians and operational experts also speak a different language. Engineers could spend a lot of time explaining the process so a data scientist could crunch numbers.
Advances in technology now allow engineers to solve 80% of their daily energy concerns without the help of a data scientist. For the most challenging use cases, engineers can use industrial analytics software to collaborate with data science teams on a machine learning exercise. This empowers both groups to contribute to efficiency improvements.
The following use cases demonstrate how power plants used operational data to meet their energy management goals.
Increasing Efficiency of a Turbine
A cycle gas power plant for producing electricity had two turbines that power a generator. The primary turbine uses hot air, which is generated by burning the gas. The secondary turbine uses steam that is generated by heating up water with the air coming from the primary turbine.
Over time, the performance of a unit of the plant’s power station began to worsen. This led to gradual losses in capacity and revenue. Because they were gradual, the performance issues went unnoticed at first. After a few years, however, the issues became more apparent. Engineers wanted to quantify the effect and duration of their losses.
Performance of the power station depends on the ambient temperature, so engineers started by finding periods where the ambient temperature was relatively constant. A value-based search was performed to find periods where the ambient temperature was between 26-28 degrees Celsius.
They also filtered out periods where the plant was taken offline or when demand for power generation was low. Once this was completed, the average power production could be calculated and exported for assessment. The results showed that the performance of the current power station decreased about 2% over the course of three years.
Engineers then compared layers of good performance periods from 2014 and bad periods from 2015-2016. They found that the gas fuel flow, the compressor discharge temperature, and the inlet guide vanes reference angle differ consistently for both groups of layers.
The resulting observations lead to the hypothesis that the root cause of the problem could be explained because of non-calibrated inlet guide vanes, which impeded air and fuel supply and ultimately power generation in the gas turbine.
After years of reduced power production and tangible revenue losses, plant engineers were able to efficiently identify and resolve the capacity loss in less than four hours by leveraging operational data. The improvement potential was about 2 Megawatts of extra production, resulting in an estimated $260,000 per year in business value.
Monitoring for Pollution on a Shaft
A combined cycle coal power plant uses an axial shaft to drive a turbine. Over time, the shaft becomes polluted and therefore less effective. In normal conditions, the axial shaft position is +/- 0.75mm. When the position exceeds a certain limit, the process shuts down and the turbine stops.
Engineers needed to find the right time to clean the axial shaft so that it did not cause undesired turbine shutdowns. They analysed the evolution of the axial shaft position and estimated when it might lead to a shutdown. The dataset they chose contained a period of several months. The axial shaft’s position can be identified by a signal with a lot of noise.
Using this information, they created a tag by uploading a csv file that simulates an x-axis with a scale that started at zero. This newly created tag then was used along with process influence factors to determine the linear equation of the shaft displacement.
As the shaft displaces further, it gets closer to needing preventive maintenance. The result is a tag that lets engineers know how many days they have until the axial shaft position reaches the allowed limit.
With information provided in operational data, engineers now know precisely how to predict when the axial shaft will need to be cleaned. They will avoid plant shutdowns and major problems with the turbine.
Conclusion
The concept of energy management is not new, but the race to Net Zero operations has again brought it to light. By leveraging operational data, energy companies can find ways to monitor, analyse, and predict energy use in power generation.
The clues hidden in operational data provide details about process behaviour over time. Today, engineers can access these details themselves to make data-driven decisions about energy management. The result is improved performance of power plant equipment, lower cost of operations, and a reduced carbon footprint.
With advanced industrial analytics, engineers can help their companies reach new milestones of energy efficiency and savings.
