Biomass Powerplant

Overview

The methods used to evaluate the operational efficiency of biomass power plants that generate energy by burning wood chips include Thermal Efficiency, Fuel Utilization Efficiency, Electric Conversion Efficiency, and Overall System Efficiency. However, these methods have limitations in adequately reflecting the fuel characteristics, equipment conditions, and operating conditions.

To overcome these limitations, SDT has developed a technology that can accurately measure the fuel, equipment, and operating conditions in real-time. This technology allows for a more reliable analysis of the power plant's efficiency and is designed to operate without interruption or manual intervention through measurements based on cameras installed inside the equipment and LIBS (Laser-Induced Breakdown Spectroscopy), and it is currently being used in Asia's largest biomass power plant.

The problems faced by biomass power plants

To operate biomass power plants at optimal efficiency, it is essential to accurately measure and continuously track the particle size and the chemical composition of the feedstock, which is wood chips. If the particle size distribution and chemical composition are inappropriate, phenomena such as agglomeration, fouling, and slagging may occur, leading to equipment corrosion and operational disturbances.

Currently, in most biomass power plants, operators collect samples of wood chips and bottom ash 1 to 2 times a day and transport them to the laboratory for analysis. However, laboratory analysis and report preparation take about two days, which limits the reliability of the analysis results and real-time response. In other words, often the cause of the problem is already affecting the equipment before the analysis is completed.

Solution of SDT -
Particle size analysis using machine vision

The wood chips and bottom ash transported to the laboratory for analysis are a tiny fraction of the total input amount. There are many limitations in relying on such small samples to predict power generation. To address these issues and optimize the operation of biomass power plants, SDT has introduced an innovative method that analyzes particle size and chemical composition in real-time on a conveyor belt, rather than in a laboratory.

SDT has installed a camera on a drag chain to monitor the agglomeration phenomenon occurring during the biomass power generation process in real-time. Since the drag chain operates in a harsh environment where the internal temperature exceeds 100°C and there is a possibility of collision with large foreign substances, we provided a design that safely protects machine vision equipment like cameras all at once. Additionally, we maintained optimal shooting conditions by freely adjusting the aperture and focus according to the optical environment. To achieve this, SDT installed a cabinet on the drag chain and safely placed equipment, including cameras, lights, air coolers, and foreign object avoidance devices, inside it.

The installed camera captures images of the bottom ash approximately once every second, and the captured images are immediately uploaded to the database. SDT's machine vision algorithm performs segmentation on the uploaded images for the particles and generates a particle size histogram through the process of measuring the diameter of each particle and counting them.

SDT's Solution -
Real-time Element Analysis on the Fairway

SDT has introduced LIBS (Laser-Induced Breakdown Spectroscopy) technology to precisely analyze the chemical composition of wood chips. This LIBS solution consists of a laser, a spectrometer, NodeQ, which is SDT's data collection module, and an industrial computer, the ECN.

The LIBS equipment uses a powerful laser to irradiate the wood chips, and through this process, analyzes the unique wavelengths of light emitted by the atoms in the wood chips after they absorb the laser energy and emit it a few nanoseconds later. To conduct bulk analysis of the elemental composition of wood chips in real time, SDT has installed a vacuum suction device on the conveyor belt. This device sucks in the wood chip samples passing on the conveyor belt, performs LIBS analysis, and then returns them to the conveyor belt.

SDT's data collection module, NodeQ, reads the wavelength values of this light in real time and transmits them to the ECN. Based on the received data, the ECN immediately identifies the elements corresponding to each wavelength through machine learning, systematically stores this information, and alerts the administrator if any anomalous elements are detected.