A METHOD AND A SYSTEM FOR MONITORING AND ON-LINE DETERMINING OF A CALORIFIC VALUE OF SOLID FUEL THAT IS CURRENTLY COMBUSTED IN A BOILER

Abstract

A method for online monitoring and determining the calorific value of a solid fuel that is currently combusted in a boiler, that includes: on-line measuring the operational data of the boiler and of at least one mill during the operation 10 of the boiler; collecting the historical data; calculating the energy balances of the steam production system; iteratively determining the efficiency of the boiler by: determining sets of mill characteristics, the fuel mass flux, and the actual calorific value of the fuel for the historical data; training a model based on artificial intelligence algorithms to predict the calorific value using the historical data and measured operational data; determining in real time, using the trained model, the calorific value of the solid fuel that is currently combusted.

Claims

1. A method for on-line monitoring and determining a calorific value of a solid fuel that is currently combusted in a boiler, wherein the boiler comprises a combustion chamber with a steam production system, to which the solid fuel is fed from at least one mill connected to a hopper, the method comprising the following steps: measuring, on-line during the operation of the boiler, operational data of the boiler, and operational data of at least one mill; wherein the operational data of the boiler are measured by means of sensors and comprise at least one of: temperature, pressure, steam flow; wherein the operational data of the at least one mill are measured by means of sensors and comprise at least one of: mill power, air pressure upstream of the mill, solid fuel feeder revolutions, air temperature upstream of the mill, temperature of the dust-air mixture downstream of the mill; collecting historical data of measurements of the calorific value of the solid fuel and the operational data of a power block, wherein the operational data of the power block include at least one of: lower heating value, mill characteristics, fuel flow rate and boiler efficiency, operational parameters of turbine, including flows, pressures and temperatures for live steam, superheated steam and secondary steam, operational data of the boiler, operational data of at least one mill, composition of exhaust gases, power of the block; calculating energy balances of the steam production system, based on the collected historical data and depending on the amount of thermal energy input to the turbine generator set divided by the boiler efficiency; iteratively determining an efficiency of the boiler based on the historical data by: determining sets of mill characteristics depending on a calorific value of the solid fuel and operational data of the mill while combusting that solid fuel, wherein the mill characteristics include at least one operational parameter selected from the group consisting of: mill power, air pressure upstream of the mill, solid fuel feeder revolutions, air temperature upstream of the mill, temperature of the dust-air mixture downstream of the mill; determining a fuel mass flux based on the set of mill characteristics; and determining the actual calorific value of the fuel for the historical data; training a model based on artificial intelligence algorithms to predict the calorific value using the historical data of fuel calorific value and measured operational data of the boiler; and determining in real time, using the trained model, the calorific value of the solid fuel that is currently combusted.

2. The method according to claim 1, wherein the historical data of the calorific value of the solid fuel are determined by means of laboratory measurements carried out cyclically or by measurement systems operating continuously, for the solid fuel prior to putting the solid fuel to the hopper.

3. The method according to claim 1, further comprising measuring, on-line during the operation of the boiler, data of ambient conditions, wherein the data of ambient conditions are measured by means of a sensor and comprise at least one of: ambient air temperature, ambient air pressure, ambient air humidity.

4. A system for on-line monitoring and determining a calorific value of a solid fuel that is currently combusted in a boiler that comprises a combustion chamber with a steam production system to which the solid fuel is fed from at least one mill connected to a hopper, the system comprising: an interface to sensors for measuring operational data of the boiler and operational data of at least one mill; wherein the operational data of the boiler comprise at least one of: temperature, pressure, steam flow; and wherein the operational data of the at least one mill comprise at least one of: mill power, air pressure upstream of the mill, solid fuel feeder revolutions, air temperature upstream of the mill, temperature of the dust-air mixture downstream of the mill; an archive module for collecting historical data of solid fuel calorific value measurements and operational data of the power block, wherein the operational data of the power block include at least one of: lower heating value, mill characteristics, fuel flow rate and boiler efficiency, operational parameters of turbine, including flows, pressures and temperatures for live steam, superheated steam and secondary steam, operational data of the boiler, operational data of at least one mill, composition of exhaust gases, power of the block; a module for calculating energy balances of the steam production system on the basis of the historical data collected in the archive module and depending on the amount of thermal energy input to the turbine generator set divided by the boiler efficiency; a module for iterative determination of boiler efficiency on the basis of historical data, configured to: determine sets of mill characteristics depending on a calorific value of the solid fuel and operational data of the mill while combusting that solid fuel, wherein the mill characteristics include at least one operational parameter selected from the group consisting of: mill power, air pressure upstream of the mill, solid fuel feeder revolutions, air temperature upstream of the mill, temperature of the dust-air mixture downstream of the mill; determine the fuel mass flux on the basis of the set of mill characteristics; and determine the actual calorific value of the fuel for historical data; a model based on artificial intelligence algorithms for predicting calorific value, that is trained using historical data of fuel calorific value and measured operational data of the boiler, and once trained, configured to determine in real time the calorific value of the solid fuel that is currently combusted on the basis of the currently measured operational data of the boiler.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0026] The object of the invention is shown by means of an embodiment in the drawing, wherein:

[0027] FIG. 1 shows a simplified diagram of the boiler;

[0028] FIG. 2 shows a flowchart of the method according to the invention;

[0029] FIG. 3 shows an example set of mill characteristics;

[0030] FIG. 4 shows a simplified diagram of the module for on-line determination of calorific value;

[0031] FIG. 5 shows a simplified flowchart of the method for determining the calorific value.

DETAILED DESCRIPTION OF EMBODIMENT

[0032] The embodiment shown herein will be discussed based on an example of a boiler supplied with a solid fuel that is coal, but the method can be used in an equivalent manner for any other solid fuel, such as biomass.

[0033] FIG. 1 shows a simplified diagram of the boiler. This diagram is based on known constructions, provided for exemplary purposes only, to illustrate an example of the system for which the method according to the invention can be used. Therefore, it is possible to apply the method also for other boilers having a similar configuration. The boiler comprises a combustion chamber 10 with a steam production system. The combustion chamber 10 comprises burners 11 to which solid fuel from mills 13 is supplied via dust-air ducts 12. Each mill 13 has its own mill hopper 14 with a solid fuel feeder that feeds the fuel to the mill. The mills 13 mill the solid fuel and separate undesirable components. Operational parameters of the mills 13 are measured by sensors 13-1 and comprise at least one of: mill power, air pressure upstream of the mill, solid fuel feeder revolutions, air temperature upstream of the mill, temperature of the dust-air mixture downstream of the mill.

[0034] Furthermore, ambient conditions are measured by at least one sensor 17-1 (i.e. a single sensor located at a designated point within the facility or a plurality of sensors located at specific locations) and comprise at least one of: ambient air temperature, ambient air pressure, ambient air humidity, i.e. data on atmospheric conditions.

[0035] A boiler drum 20 separates water from steam. Live steam having temperature T1, pressure P1 and flow F1 (measured, for example, as flow rate) is directed to a live steam superheater 30, wherein additional energy is supplied to the steam by raising its temperature and thus superheated steam having temperature T2, pressure P2 and flow F2 is fed to a high-pressure section 41 of a turbine, wherein it is subject to expansion, releases some of its energy and returns to the boiler as secondary steam having temperature T3, pressure P3 and flow F3. A secondary steam superheater 40 heats the steam and the superheated steam having temperature T4, pressure P4 and flow F4 is directed to a medium-pressure and then to low-pressure section 42 of the turbine, wherein it releases its energy. The values of temperature, pressure and flow of the steam are measured by means of corresponding sensors 15-1, 15-2, 15-3, 15-4.

[0036] FIG. 2 shows a flowchart for carrying out the method according to the invention.

[0037] In step 201, the boiler useful thermal power (Qin) is determined, which indicates the amount of energy transferred from the boiler 10 to the turbine 41, 42 in the form of steam. At this step, the calorific value of the fuel and the fuel mass flow rate are not known yet, therefore it is not possible to calculate the efficiency from the energy output of the boiler. The boiler useful thermal power (Qin) is determined based on the data on the flow of live steam, secondary steam, the parameters of this steam and other parameters, for example using the formula:

[00001]$Q_{in}=\left[\left({\stackrel{.}{m}}_{p\stackrel{}{s}w}-{\stackrel{.}{m}}_{\mathrm{wtrysk}\_p\stackrel{}{s}w}\right)*\left(h_{p\stackrel{}{s}w}-h_{\mathrm{wz}}\right)+{\stackrel{.}{m}}_{\mathrm{wtrysk}\_p\stackrel{}{s}w}*\left(h_{p\stackrel{}{s}w}-h_{\mathrm{wtrysk}\_p\stackrel{}{s}w}\right)+\left({\stackrel{.}{m}}_{\mathrm{pwt}}-{\stackrel{.}{m}}_{\mathrm{wtrysk}\_\mathrm{pwt}}\right)*\left(h_{\mathrm{pwt}}-h_{p\stackrel{}{s}w\_\mathrm{wylot}}\right)+{\stackrel{.}{m}}_{\mathrm{wtrysk}\_\mathrm{pwt}}*\left(h_{\mathrm{pwt}}-h_{\mathrm{wtrysk}\_\mathrm{pwt}}\right)\right]/3600-Q_z$

wherein: [0038] {dot over (m)}.sub.p?w is the flow of live steam (F2) entering the turbine (41) [0039] {dot over (m)}.sub.wtrysk_p?w is the flow of water injected into live steam [0040] h.sub.p?w is the enthalpy of live steam entering the turbine [0041] h.sub.wz is the enthalpy of water fed to the boiler [0042] h.sub.p?w_wylot is the enthalpy of live steam leaving the turbine [0043] h.sub.wtrysk_p?w is the enthalpy of water injected into live steam [0044] {dot over (m)}.sub.pwt is the flow of secondary steam (F4) entering the turbine (42) [0045] {dot over (m)}.sub.wtrysk_pwt is the flow of water injected into secondary steam [0046] h.sub.pwt is the enthalpy of secondary steam entering the turbine [0047] h.sub.wtrysk_pwt is the enthalpy of water injected into secondary steam [0048] Q.sub.z is the amount of heat dissipated outside the system

[0049] In step 202, the initial efficiency of the boiler is determined on the basis of historical 24-houraverage laboratory data of calorific value and the composition of the fuel, according to the PN-EN 12952-15:2004 standard.

[0050] In step 203, processing of measurement data from the environment of the coal mills is carried out so as to remove erroneous indication values generated by currently deactivated mills 13, e.g. the error of negative revolutions of feeders of a non-operating mill 13.

[0051] In step 204, it is initially assumed that mill characteristics are identical for all of the mills 13, which corresponds to the assumption that each mill hopper 14 contains a fuel with the same properties.

[0052] Based on the data from steps 202, 203, 204, in step 205 the fuel mass flow rate is adjusted by selecting appropriate mill characteristics. The resulting fuel mass flux is related to the 24-hour-average laboratory calorific value, so that the balance of the amount of chemical energy of the fuel fed to the boiler is fulfilled. The balance relates to the amount of thermal energy input to the turbine generator set (Qin), divided by the boiler efficiency, according to the formula:

[00002]$\mathrm{qPal}=f\left(.Math.\mathrm{obrMW}\right)*{\mathrm{LHV}}_{\mathrm{lab}}=\frac{Q_{in}}{{?}_{\mathrm{boiler}}}$

wherein: [0053] ?obr MW is the measured sum of the feeders revolutions [0054] ?(?obrMW) is the fuel mass flow rate as a function of the sum of revolutions of the feeders of particular mills, and the calculations are carried out using the determined mill characteristics [0055] LHV.sub.lab is the 24-hour-average laboratory calorific value [0056] Qin is the amount of thermal energy input to the turbine generator set [0057] ?.sub.boiler is the boiler efficiency

[0058] In step 207, for the steady states of boiler operation provided in step 206, linear mill characteristics 1-4 are created as shown in FIG. 3 (determining the dependence of the mass flow [expressed in t/h] on the load of the mill feeder motors (which is translated to coal feeder revolutions) [expressed in %]) based on determination coefficient metric R2.

[0059] As a result, in step 208, a set of mill characteristics is obtained, i.e. individual characteristics for different fuel qualities, which are a measure of the coal's susceptibility to milling, as for example shown in FIG. 3.

[0060] In step 209, the current mill characteristics are detected (individually for each mill) based on the measured operational parameters of the coal mill, such as mill motor power, pressure of air entering the mill, revolutions of mill feeders.

[0061] On the basis of a particular current mill characteristics, the mass flow of fuel fed by a particular mill is determined in step 210. The characteristic is selected depending on at least one currently measurement parameter of the mill (when the measured parameter is changed, another characteristic is selected).

[0062] Then, in step 211, the estimated LHV of the fuel is determined based on the energy balances and the determined boiler efficiency and it is reported as the result of the procedure.

[00003]$\mathrm{LHV}=\frac{{\stackrel{.}{Q}}_{in}}{{\stackrel{.}{m}}_{\mathrm{pal}}*{?}_{\mathrm{boiler}}}$

wherein {dot over (m)}.sub.pal is the mass flow of fuel

[0063] In step 212, the boiler efficiency for the thus estimated LHV is determined according to the PN-EN 12952-15:2004 standard.

[0064] The boiler efficiency is determined on an ongoing basis (on-line) for a particular historical moment, e. g. for every minute. After updating the calorific value, the boiler efficiency is calculated again.

[0065] The LHV estimated in step 211 and the boiler efficiency estimated in step 212 are provided as input to step 207, allowing a more accurate determination of the set of characteristics.

[0066] Thus, steps 207-212 are performed iteratively until the boiler efficiency calculation error (i. e. the difference between successive iterations of steps 207-212) converges to a value below a predetermined threshold, for example of the order of 10.sup.?3. The calculation error is defined as |?.sub.boiler.sub.n-1??.sub.boiler.sub.n|<10.sup.?3.

[0067] The process described in FIG. 2 in steps 207-212 is implemented on historical data of the block operation (i.e. data collected over a certain period of time, for example several months or several years) and on laboratory data of fuel properties. The historical data and results prepared in this way, in particular the estimated LHV, mill characteristics, fuel flow rate and boiler efficiency (after obtaining the error convergence) are used to train and test an artificial intelligence model configured to predict the current calorific value of the fuel. Other operational data of the block can be used as well, such as operational parameters of turbine, including flows, pressures and temperatures for live steam, superheated steam and secondary steam, operational data of the boiler, operational data of at least one mill, composition of exhaust gases, power of the block.

[0068] FIG. 4 shows a simplified diagram of the module for on-line determination of calorific value. The data are retrieved from a database 401 (or other type of archive module), both for preparation and on-line operation of the model. In module 402, the data are validated (by removing incorrect indications or by simulating their correct values and processed (e.g. by aggregation or by extracting averages from several measurement sensors) in order to obtain meaningful values for further modules. Based on the processed data in module 403, basic operational factors are determined, such as secondary steam flux (if not measured), boiler thermal power Q.sub.in, chemical energy of fuel supplied to the boiler q.sub.Pal, boiler losses and boiler efficiency. In module 404, operation characteristics for the coal mills (as shown in FIG. 3) are determined for steady states of boiler operation, until the error in boiler efficiency calculations is minimised and converged, according to steps 207-212. In module 405, the calorific value is recalculated for historical data which, in combination with the measurement sensor data retrieved in module 406 (i.e.: live and secondary steam parameters, mill system operational parameters, parameters of air supplied to the boiler, flue gas parameters, operating settings, flue gas denitrification system operational parameters, boiler support equipment operational parameters and atmospheric parameters), are used to train the model based on artificial intelligence algorithms 407 to determine the calorific value. The prepared model 408 predicts the current calorific value of the fuel based on the measurement data retrieved in module 406. An artificial intelligence model 408 based on recursive elements is used, which allows the effect of energy accumulation and the response delays of automation and control systems to be included in the model. The module 408 is trained by means of the module 407 such that the historical operational data of the boiler operation with the corresponding calorific value of the fuel is used for training the artificial intelligence model. Then, such trained model 408 can be used to predict on-line (i. e. on an ongoing basis) the calorific value of the fuel based on current operational data.

[0069] FIG. 5 shows a simplified diagram of the method for determining the calorific value, which is a generalisation of the flowchart in FIG. 2. In step 501, the operation characteristics of the mill 13 for different types of coal are determined. Then, in step 502, current efficiency of the boiler is determined. In step 503, current operation characteristics of the mill 13 are detected, and in step 504, calorific value of the fuel is determined on an ongoing basis (for example, for each minute).