Source: Dianlian Intelligent Manufacturing
Recommended unit: Datang Sanmenxia Power Generation Co., Ltd
The author of this article: Tao Ze
Previous
Big data will bring about a revolution in management (digital revolution), and the successful implementation of the Group's three major fuel projects can monitor all aspects of fuel management and realize the digitization of fuel management processes. Fuel big data can not only realize the monitoring and analysis of the fuel management process, but also use these fuel data to control the fuel furnace management process, which can produce a new management idea and bring a new change. Here, the utilization of fuel big data is explored, combined with the actual production process, to solve the problems existing in the current deep coal blending, and to explore the automatic control of the coal conveying system and furnace blending in the future.
1. Background and significance of the study
The era we live in is the era of big data, the era of artificial intelligence, with the development of computer and sensor technology, the automatic driving function of the car has been initially realized. For the industry, the construction of smart power plants is the direction of the development of power plants in the future, and smart power plants are to achieve a high degree of automation and intelligence in all aspects of production, relying on a control center with strong computing power to make various decisions and control according to a large amount of data generated in the production process, and reduce dependence on personnel. Smart power plants are not far away, and we are also facing a revolution in fuel management, and it is not a fantasy to see what changes will be achieved in five or ten years, and whether artificial intelligence will be able to automatically control the supply of fuel into the furnace. With the application and development of the three major projects of the group, we can monitor and use more and more fuel data, using these data, the algorithm of the corresponding control model is established on the computer, which can realize the accurate prediction of the coal quality of the furnace, and finally realize the automatic and intelligent control of the whole process of the adjustment of the fuel index to meet the power generation needs of the unit. From a technical point of view, the automatic and intelligent control of fuel into the furnace has been feasible.
Second, the current development status of deep coal blending
The management of fuel into the furnace is not only to meet the needs of the unit's power generation consumption, but also to meet the requirements of the unit's safety, environmental protection and economy. For the control of the fuel index of the thermal power plant, it is necessary not only to meet the needs of the safe and environmentally friendly operation of the unit, but also to appropriately reduce the coal quality of the furnace according to the adaptability of the unit to the fuel index, so as to reduce the cost of the fuel entering the furnace.
At present, coal is graded by quality, and the price is judged by quality. Coal blending is based on the condition that the main equipment and environmental protection equipment have a certain margin under medium and low load conditions, and under the premise of meeting safety and environmental protection, the ratio of different coal types is adjusted to achieve the furnace index in a suitable range and purposefully reduce the indicators related to coal prices. According to the actual situation, the general adjustment of the coal index includes five indicators: received base low calorific value, dry ash-free base volatile, received base total sulfur, and received base ash. Among them, calorific value, sulfur and coal price are the most sensitive, and are often used as the main adjustment indicators of coal blending. We carry out coal blending to achieve this purpose, coal blending is an important means to stabilize coal quality and increase fuel economy in thermal power plants under the current conditions.
Therefore, the control of coal index is the core of fuel furnace management, the fuel consumption process of the boiler is continuous, and the blending is the process of mixing different coal types in the process of fuel entering the furnace to achieve the expected index range. The blending methods are divided into static blending and dynamic blending. Static blending is that the fuel is mixed in advance, such as mixing by a coal pusher in the coal yard before use, and dynamic blending is directly used after blending in the fuel transportation process through the ring coal feeder and belt.
(1) The current process of coal blending
Our blending work is to purposefully blend the fuel through the main nodes of the coal conveying system during the process of feeding the fuel into the boiler. Taking the specific process of blending in Sanmenxia Power Plant as an example, the main nodes of the coal conveying system are coal unloading ditch, coal yard, bucket turbine, coal conveying belt, raw coal silo, ring coal feeder and coal mill raw coal bunker, and the four times of "primary, coarse, fine and fine" are implemented. That is, the initial blending during unloading, the mixing and coarse mixing before the silo, the fine mixing according to the coal quality of the raw coal bunker after the silo, and the fine blending in the furnace is realized by adjusting different grinding forces. Automobile coal is divided according to the sulfur content of different sulfur, calorific value and volatile content in the automobile coal unloading ditch, and is received and unloaded separately, and stored in the coal yard according to the main control indicators of blending. The second crude blending of the same coal type after the coal type arrives at the plant is carried out by using the coal yard stratification and bucket wheel to improve the uniformity and doping of coal entering the cylindrical silo. According to the coal quality data of the fuel silo, the coal conveying professional uses the blended coal data template for blending planning. At the same time, by increasing the number of coal loading in the raw coal bunker of the boiler, the amount of coal is controlled according to the output of the boiler coal feeder to meet the coal quality requirements of the unit in different periods.
02 Implementation of coal blending
The first is to formulate the blending plan and issue the blending plan, the value is based on the load curve of the unit before the dispatch and issuance, combined with the blending boundary under different loads, and considering the safety and environmental protection needs of equipment operation (mainly considering whether there is main equipment and environmental protection equipment decommissioning), and issue the coal quality indicators of the unit into the furnace at different load periods tomorrow.
The control value of coal quality index of the unit entering the furnace in different load periods
According to the coal quality of the coal stored in the coal yard, the next day's coal plan and the coal quality, combined with the needs of the coal yard, the coal mixing supervisor of the coal conveying professional coal blending supervisor formulates the specific plan for coal pick-up and unloading, coal taking from the coal yard, belt blending, and coal mill blending, and issues the blending instructions for the coal conveying equipment.
According to the blending plan and blending instructions, the attendant implements the blending of the coal unloading ditch, the coal taking of the bucket turbine, the adjustment of the coal proportion of the ring coal feeder of the raw coal silo and the adjustment of the coal distribution of the coal mill, and at the same time corrects and adjusts the blending instructions according to the deviation between the actual load change and the blending result on the day.
3. Problems in the process of coal blending:
It can be seen from the entire implementation process of the above-mentioned coal blending that we can formulate a good plan, know what kind of coal should be burned under what load and when, and can formulate a perfect specific coal blending plan and implementation measures, and can know what kind of coal can be blended and what equipment can be blended to achieve the expected coal quality target. But all this is only a plan on paper, and the specific implementation and implementation, the implementation of coal blending is no less difficult than the pure manual adjustment of the water level of the steam drum. First, the coal species to be blended in the coal blending are all imaginary coal types, and the coal quality index is an empirical estimate, which is the expected coal quality index obtained by blending in a fixed proportion, and the index of the blended coal used itself is fluctuating. Therefore, in practice, we can accurately grasp the blending ratio of different coal types and accurately control the amount of blended coal, but it is difficult for us to grasp the coal quality index of the blended coal type, due to the unevenness of coal quality, the blending results often fluctuate with the fluctuation of the blended coal index. Second, the coal conveying process has a great delay, the intermediate reserves of the coal conveying system are very large, from the raw coal into the raw coal silo to the furnace, it takes 4-6 hours under normal circumstances, and it takes 12-24 hours from the silo to the raw coal bunker, how to ensure that the required coal quality index is burned on time and at the right load. Because of such a large system delay, the amount of coal stored in the middle cannot be accurately grasped, and it is difficult to match the coal quality index with the load. In fact, the intermediate reserves of each part of the system are estimated by manually looking at the position, and the specific time when the blended coal can be burned depends on manual estimation, and the actual error is often in a few hours.
In the above figure, it can be seen that there is a lag between the sulfur and load matching curves in the operation of the unit, and the sulfur content does not decrease during multiple high-load periods. Therefore, it is not enough to analyze the quality of coal blending work only from the weighted value of coal quality of procurement or furnace caliber, but also to consider the accuracy of coal blending and the matching degree of coal quality and load of furnace blending. From the actual situation, due to the fluctuation of coal quality caused by very large safety and environmental protection risks, such as sulfur fluctuations caused by the instantaneous exceedance of sulfur dioxide, smoke and dust instantaneous exceedance, in the current environmental protection high-pressure situation, are very big risks. Therefore, the accuracy of coal blending is insufficient, the accuracy of coal blending cannot be controlled, the index of furnace entry fluctuates, and a large safety and environmental protection margin is forced to be left during blending, which seriously restricts the development of deep coal blending.
Fourth, the cause of the above problems analysis
First, the raw coal mixing is limited by the blending method, and it is difficult to mix evenly. Since the raw coal is a bulk cargo, its solid particles cannot be mixed spontaneously, and the blending process must rely on a fixed process through the blending of fixed equipment. Taking the Sanmenxia Power Plant as an example, during the period of heavy load on the power grid, the daily consumption of raw coal into the furnace is more than 20,000 tons. The natural bulk density of 20,000 tons of raw coal is calculated to be more than 20,000 cubic meters, and it is obviously impossible to stack such a large volume of coal in the coal yard and use a coal pusher to push the coal back and forth in the coal yard for mixing. Therefore, the coal blending is mainly through the belt this link, so that different coal types are mixed, and the mixing process is not to pour ink in the vat, but to twist the rope like a rope, to blend different coal types together, and its mixing characteristics determine that if the coal quality of the blended coal type fluctuates, the coal quality fluctuation will be directly transmitted to the blending result and enter the next process.
Second, it is difficult to popularize the online measurement technology of coal quality, and the coal quality index has not been measured in real time. If the coal blending is compared to a control system, the coal quality index, the amount of coal blending is the input, we can accurately control the proportion of different types of coal mixing (coal volume), we can accurately control the frequency and coal feed of the raw coal silo ring coal feeder, but for the core index of coal quality, we can not accurately grasp it in real time. Unlike parameters such as pressure, temperature, velocity, and PH value, coal quality indicators are easy to measure instantly. At present, the on-line monitoring and analysis technology of coal quality includes X-ray fluorescence technology, neutron-induced instantaneous γ-ray analysis technology and dual-energy γ-ray projection technology, but X-ray fluorescence technology is only suitable for measuring elements with atomic number greater than 11, and the measurement accuracy and sensitivity are not high γ. At present, there are few applications of online coal quality monitoring in coal-fired power plants in China, and with the stricter management of major hazards, few power plants have used radioactive sources for online coal measurement considering the hazards of radioactive sources and the accuracy of measurement. According to the actual situation of the power plant, the coal quality test index is generally manually tested, and it takes 12-24 hours to get the results, which is far from meeting the needs of the coal blending index for timeliness, so we can only make a rough estimate based on historical data.
Third, the actual coal source of the power plant is complex, and the result of coal blending is often greatly affected by the coal source structure. For power plants along the coastal rivers, there are few types of coal in the water, the coal quality is relatively stable, and it is easy to blend coal. For most inland power plants, the source of thermal coal is unstable, the type of coal is varied, and the quality of coal is uneven, and it is difficult to change this situation in the short term. The maximum number of coal mines in Sanmenxia Power Plant reached 34 in a single month, the calorific value of incoming coal at low level increased from 2,000 kcal/kg to 6,300 kcal/kg, and the sulfur content of incoming coal increased from 0.3% to 3.3%. So many types of coal can be classified according to calorific value and sulfur can be divided into low heat and high sulfur, high heat and high sulfur, low heat and low sulfur, medium heat and high sulfur, medium heat and medium sulfur, etc., if the volatile matter, ash melting point and other indicators are included in the classification, there are more types. In the face of so many coal types, it is not possible to store all coal types separately, but to store them separately as needed. If the coal quality is mixed and stored, even if the coal is mixed and then burned, the coal quality stored in the coal yard will change due to the structure of the coal, and the coal quality of the first storage and the later storage are different.
Fourth, due to the restrictions of the coal conveying system, part of the incoming coal must be directly burned in the furnace and cannot be burned through the coal yard. Taking the Sanmenxia Power Plant as an example, during the high-load period in summer and winter, the daily amount of coal entering the plant and furnace is 2.3-25,000 tons. Due to system limitations, if coal is taken from the coal yard, the bucket turbine will not be able to pile coal, and the coal receiving and unloading must be stopped. Such a large boiler consumes coal, and it is not allowed to pile it all into the coal yard and then burn it for use. Since the coal is immediately put into the furnace, there is a problem, the problem of the rhythm of the coal. For the same day, the weighted average coal quality is fine, but the different rhythms will cause the coal quality to fluctuate greatly.
Fifth, the coal quality of some ore sites itself fluctuates. For raw coal, it will fluctuate due to the crossing section of the excavation face and the abnormality of the washing equipment, but the general fluctuation is not large, but the intermediate supplier will supply coal, and some of it will be blended before entering the factory, and its coal quality index will fluctuate.
Sixth, the silos and raw coal bunkers of the coal conveying system have large intermediate reserves, which are difficult to accurately quantify and calculate. From the perspective of automatic control, coal conveying is a continuous control system with large delay characteristics. Even if the quality of the blended coal is precisely controlled, because the system has a large storage capacity, the blended coal is often burned after several hours, which has a great delay. For example, we use anti-aircraft guns to hit planes, but it takes 4-6 hours for our shells to fly to the height of the plane, and such a large amount of delay requires me to predict when to launch, so it is very important to calculate the timing of blending and coaling. At present, with the characteristics of the power grid, the peak regulation of the unit is becoming more and more frequent, and the load curve changes greatly with the power grid situation. Because it is difficult to accurately control and calculate the amount of cached coal in the unit system, the coal quality of the unit is mismatched with the load or vinegar, resulting in the occurrence of load limiting and environmental protection events in the unit. Therefore, coal blending is two-dimensional, with time characteristics, and the right coal quality must be added at the right time to achieve the desired effect.
The principle of coal blending seems to be simple, but the system delay cannot be accurately grasped because the fluctuation of blended coal quality cannot be measured. In practice, the raw coal bunker can only be replenished with a fixed blending ratio, estimated coal quality, and approximate time, so the coal quality results often deviate greatly from expectations.
5. Exploration of using the big data of the three major fuel projects to solve the problem of coal blending
According to the problems existing in the current deep coal blending work, solutions are found from big data. In 2016, the "three major projects" of the group company were fully completed in the group, opening a new era of digital fuel for the group company. The "Three Major Fuel Projects" renovate the metering, mining, coal yard and other related equipment to collect the original data of quantity, quality and price corresponding to fuel receipt, consumption and storage in a real-time, accurate and reliable manner. In terms of incoming inspection, it realizes the functions of entry registration, intelligent queuing, measurement management, mechanical sampling, sample bonding, gangue deduction and tonnage, factory registration, etc.; in the digital coal yard system, it has developed the functions of coal yard partitioning, stacking and taking instructions, warehousing, delivery, inventory, three-dimensional dynamic display, and collection and consumption statistics, and collected real-time coal yard equipment and coal conveying program control data, so as to realize the linkage control and refined management of software and hardware in the digital coal yard; Laboratory work is standardized, digitized, and automated.
The first is to realize the big data analysis of coal quality at each mining site.
Through the analysis of the data of the three major projects, the problem of large fluctuations in coal quality at a single mine site can be found in time, and the impact of coal quality fluctuations at the incoming and outgoing mines can be analyzed. According to the sulfur analysis of the coal quality of 33 suppliers purchased by Sanmenxia in July 2018, 15 suppliers had sulfur fluctuations of less than 0.1%, 12 suppliers had sulfur fluctuations of 0.1-0.2%, 4 suppliers had sulfur fluctuations of 0.2-0.5%, and 2 suppliers had sulfur fluctuations of more than 0.5%, and it was found that the mines with sulfur fluctuations of less than 0.2% accounted for 85% of the coal intake By analyzing the fluctuation of coal quality at each mine site one by one, it is concluded that the coal quality of each mine is relatively stable, which is not the main reason for the fluctuation of coal quality index after blending.
The second is to achieve accurate prediction of the change of coal quality in the plant.
In practice, the production process is continuous, and the accurate prediction of the coal quality index is meaningless if the departure time index. Therefore, the process of receiving and unloading coal-fired into the plant and the process of coal-fired blending are continuous in time, and they are two-dimensional parameters with a time dimension. For inbound coal, from the first car of coal every day to start receiving and unloading, the ore points and unloading structures of the unloading per hour are different, resulting in the quality of the coal entering the plant per unit hour, and the incoming indicators change over time, and these coal quality changes will be transmitted to each process after the incoming coal unloading. The number of ore sites receiving and unloading in Sanmenxia Power Plant is normally more than 15 per day, and the maximum number of ore sites is about 25. Considering the actual handling, unloading and blending, the coal quality is divided into three types based on the similarity of coal quality, which are high calorific value and low sulfur main coal, low sulfur economic coal, and high sulfur blended coal. The following takes the receiving and unloading situation of a specific mine site as an example to illustrate the change of coal quality in a certain day due to the coal inlet structure over time. Ningxin mine, Macun mine, Qinjin mine are all high heat and high sulfur steam coal, the coal quality is shown in the following table, according to the blending scheme arranged in the same coal unloading ditch for unloading, the daily transfer volume is 3000 tons, 1200 tons, 2000 tons, according to the coal quality of the mine, the daily weighted calorific value is 4990 kcal/kg, weighted sulfur 2.8%.
Due to the different road conditions of each mine to the plant, from 08:00 to 18:00 on a certain day, the amount of coal from the three mines to be unloaded is as follows, and the coal quality of the unloading ditch is calculated for hours. It is estimated that although the weighted calorific value of high-sulfur coal is 4990 kcal/kg and the weighted sulfur content is 2.8%, due to the difference in coal quality at each mine site and the different structure of entering the plant, the hourly sulfur content is the lowest 2.4% and the highest 3.3% throughout the day, with a fluctuation range of 0.9%.
In the process of coal-fired incoming and unloading, the ideal situation is that the coal structure of each mine site is as average as possible per unit hour, but it is difficult to achieve it due to factors such as road, sampling, and unloading. Especially during the peak summer season. The amount of coal to be unloaded barely meets the demand for consumption, and the automobile coal cannot be detained after arriving at the plant, and it is difficult to achieve the idealization of the rhythm of entering the plant and unloading at each mining point, and the coal quality often fluctuates when entering the plant.
The three major projects carry out comprehensive management of the incoming vehicle information, and the system includes the vehicle registration time, sampling, overweight, pick-up and unloading, and empty-back time, as well as information such as mine type, ticket weight, gross weight, net weight, and pick-up and unloading location. Through the above information, we can grasp the unloading situation per unit hour, calculate the unloading vehicles and unloading tonnage of different mining sites, and calculate the real-time coal quality of each unloading ditch according to the historical coal quality information of the mining site, so as to provide reference for the subsequent blending and adjustment work.
The third is to use the equipment and data recording functions of the three major projects such as truck scales, rail scales, belt scales, etc., to establish coal storage models such as raw coal silos, coal mill silos, and coal unloading ditches, and quantify each system of the coal conveying system according to the principle of quality conservation and balance of supply, consumption and storage. Quantification is to realize that each node of the system can calculate the amount of coal through, and can accurately track the situation of coal flow in the system, such as the change of coal quality in the unloading ditch, the impact of coal quality change on the subsequent parts, and the fluctuation of coal quality change can be accurately grasped, and when to pass through each belt, raw coal silo silo, coal feeder, coal mill, to realize the short-term forecast function of the coal quality exported by each coal storage equipment.
Taking the coal unloading ditch as an example, the coal inlet of the coal unloading ditch is calculated by the truck scale, the coal output of the coal unloading ditch is calculated by the belt scale, and the stratified coal quality of the coal in the coal unloading ditch can be measured according to the coal unloading situation of the coal unloading ditch, and the coal quality of the coal outlet of the coal unloading ditch can be measured according to the relationship between the quantity and the coal quality. Taking the raw coal bunker of the boiler mill as an example, the coal belt scale and the plough can be used to accurately calculate the amount of coal on a coal mill, and according to the consumption of the belt scale of the coal feeder, a rolling model of coal consumption in the raw coal bunker can be established, which can accurately calculate how long the current coal quality can be burned, so the coal quality of the boiler can be predicted within 1-4 hours in a short term.
The advantage of accurately establishing a coal storage model is that as long as you know the quality and rhythm of incoming coal, you can accurately grasp the coal quality and fluctuation of incoming coal in the follow-up situation, and can track the fluctuation of coal quality in the system, and make adjustments in the follow-up process.
6. The coal monitoring project of automobile transportation into the plant is applied in fuel management
In order to grasp the most real coal quality of coal unloading in the coal unloading ditch, Sanmenxia Company has established an incoming coal monitoring system, based on the data of the three major fuel projects, real-time statistics of the ore points and quantities of receiving and unloading, estimating the real-time coal quality of the incoming coal, and grasping the first-hand accurate real-time information of the coal quality of the incoming coal receiving and unloading, which is applied to coal stacking or direct blending, laying the foundation for the informatization and digitization of the entire fuel process in the future.
The Sanmenxia incoming coal monitoring project makes full use of the big data of fuel entering the plant, and the information contained mainly includes real-time information of mine information, entry, sampling, overweight, unloading and delivery. These big data are processed and analyzed and displayed in real time.
The project is characterized by computer programming, using a small amount of code to achieve the minimum cost of big data analysis and display of the factory, the whole process is automatic, real-time update, and has a strong timeliness. It is very convenient to supervise and scientifically dispatch the unloading process, so as to optimize the unloading process and improve efficiency. By accurately grasping the unloaded ore points and quantity, and cooperating with the coal quality information data of each ore site, the unloaded coal quality can be grasped in real time.
VII. Conclusions
Through the application of big data of the three major fuel projects, this project explores the solution to the existing problems in coal blending, proposes a method for predicting the fluctuation of coal quality in the incoming plant and a method for establishing a quantitative model of the coal conveying system, and realizes the prediction and tracking of the fluctuation of coal quality in the coal conveying system. Because the three major fuel projects of the group are mainly established to realize the relevant functions of fuel quality acceptance, the relevant functions in the project still need to be manually queried and downloaded from the data system of the three major projects, and calculated and analyzed in the table, which is still inconvenient in practical application. The development of the coal conveying control system must be to realize the docking with the key equipment of the coal system, realize the automatic analysis and adjustment of the blending ratio of the coal conveying control system according to the coal structure, coal rhythm and coal storage, and realize the intelligent management of the coal into the furnace, so further research and practice are needed on this basis.