RECOVERY BOILER

FIELD OF THE INVENTION The present invention relates, in general, to a system and a method for optimization of a recovery boiler process in general. More particularly, it relates to the optimization of a recovery boiler process used in pulp mills.

BACKGROUND OF THE INVENTION Recovery boiler is a major equipment in the recovery cycle of black liquor, that is formed during digester process and other pulp making processes. The black liquor contains dissolved organic compounds (from wood) and inorganic compounds (NaOH, Na ₂ S, Na ₂ CO ₃ and Na ₂ SO ₄ ). Na ₂ S and NaOH are the required chemicals for pulp making process in digester. Na ₂ CO ₃ and Na ₂ Sθ ₄ are undesirable chemical species. Recovery boiler recovers Na ₂ S from Na ₂ SO ₄ via reduction reactions at the bottom of the furnace in recovery boiler. NaOH is recovered from Na ₂ CO ₃ in a causticizing process subsequent to recovery boiler. The recovered stream from recovery boiler and causticizer contains NaOH and Na ₂ S as major species and is fed back to the digester for pulp making process. To summarize, recovery boiler is used to (i) recover inorganic cooking chemicals (ii) generate heat energy by burning the organic materials derived from the wood and (iii) burn the organic chemicals in order not to discharge them from the mill, as pollutants.

Through the recovery boiler process, the pulp mill saves on chemical cost and hence the recovery boiler (and also causticizer) increases the economic performance of the pulp mill.

The chemical recovery boiler is a major component of the liquor cycle in a pulp mill and an important key to overall mill economic performance. Several issues add to the importance and complexity of recovery boiler operations. The variations in the calorific value (Btu content) and the temperature of black liquor, the size of the black liquor droplets, temperature and the distribution of combustion air can result in the varied performance of the boiler that in turn affects the quality of the steam generated and the emissions from the boiler. The intensely coupled phenomena taking place inside the recovery boiler make it more difficult to operate the process at the optimum. The optimum performance of the boiler implies maximum reduction efficiency to recover the cooking chemicals, reduced emissions from the boiler and maintaining the steam quality at the desired level. Thus, there is a strong need to develop a method that considers all the above factors to ensure the optimum performance of the boiler.

A modern day pulp mill uses sophisticated control systems, for example, a Distributed Control System (DCS), to regulate and optimize various processes related to the pulp mill. The DCS may as well optimize operations and production involved in the manufacture of pulp and paper. The control and optimization strategies are usually based on modeling and simulation modules available with the DCS. A variety of process models are available in the literature including that for recovery boiler and mainly new modules based on these models are made available as software/ hardware (mostly as software) solutions for the DCS positioned for pulp and paper mills.

The process models for recovery boiler in the literature are either the first principle based models or data driven models. These models are used for off-line simulations and, for control and optimization applications. The models used for control and optimization applications need to be mathematically simple enough to guarantee convergence and at the same time mathematically complex enough to capture the important dynamics and relationships among the required process variables. The data driven models have the advantage that they are mathematically simple but the use of these models is restricted to a narrow operating region of the recovery boiler. The first principle based model have the advantage that they capture the physics of the problem and thus their model predictions are reliable for a wide range of operating conditions of the recovery boiler. Hence the first principle based models are desired if they can be formulated and utilized.

A major challenge lies in the control of reduction process inside the recovery boiler. Direct measurements of chemicals, NaOH, Na ₂ S, Na ₂ CO ₃ and Na ₂ SO ₄ are not available in the recovery boiler. Therefore, it is difficult to get a direct estimate of reduction efficiency in the recovery boiler that depends on the concentration of the above mentioned chemicals. The data driven models use an indirect measure of the reduction efficiency i.e. the temperature of the char bed at the bottom of the furnace in the recovery boiler. But, reduction efficiency also depends on several other factors such as the availability of oxygen at the bottom of the bed, the chemical composition of the black liquor entering the furnace of recovery boiler, drying, volatilization and combustion reactions that take place during the flight of black liquor from the liquor nozzles to the char bed. All these phenomena mentioned above affect the concentration of the chemical species (carbon, NaOH, Na ₂ S, Na ₂ Cθ ₃ and Na ₂ SO ₄ ) that reach the char bed and hence affect the reduction reaction rate. Therefore, data driven models, that assume char bed temperature as the measure of reduction process, may be inaccurate for control of reduction process.

The inefficiency of the data driven models to accurately predict the reduction reaction rate and its inability to capture the dependence of reduction process on the various process variables of the recovery boiler result in inefficient control and optimization of the recovery boiler process.

As various coupled phenomena, such as combustion, char volatilization and reduction occur in the recovery boiler, the data driven models are not able to efficiently control and optimize the recovery boiler process. This results in poor reduction efficiency, poor combustion and higher emission of the pollutants from the recovery boiler. All these factors add to the cost of the pulp mill.

Thus, there is a strong need to develop a method and a system that accurately predicts the process variables, including the concentration of chemical species in the recovery boiler and also efficiently controls and optimizes the recovery boiler process.

OBJECTS OF THE INVENTION

The principal object of the present invention is prediction and optimization of overall performance of recovery boiler.

Another object of the present invention is to use a model based completely or partially on first principles, that reflects the relationship between all the manipulated and controlled variables of the recovery boiler

Still another object of the present invention is to formulate objective function to optimize parameters related to performance and operation of recovery boiler process.

Another object of the present invention is to use model to estimate and predict the recovery boiler parameters that are not available from online measurements.

Yet another object of the present invention is to use the process variables in optimization as well as in monitoring performance and operational parameters of the recovery boiler.

Yet another object of the present invention is to use the estimated and predicted boiler parameters to control or stabilize the processes downstream of the recovery boiler.

SUMMARY OF INVENTION

A method and a system for optimization of parameters for a recovery boiler in a pulp mill is provided. The method is based on use of a first principle mathematical model to estimate parameters that are otherwise not measurable to accurately control the performance of the recovery boiler. The estimated parameters and the developed models are used to optimize performance and operational parameters in a pulp mill, specifically that of the recovery boiler. An aspect of the present invention relates to a method for optimization of one or more boiler parameters in a recovery boiler process. The method comprises the steps of: (i) Obtaining a process model describing relationship between various process variables for one or more units of the recovery boiler system; (ii) using process variable data of one or more units of the recovery boiler system to estimate one or more boiler parameters;

(iii) using the one or more boiler parameters to develop atleast one objective function for optimization (iv) optimizing the said objective function for control of one or more boiler parameters of the recovery boiler.

Wherein, the process model for atleast one unit of the recovery boiler system is based completely or partially on a first principle mathematical model of the recovery boiler process and the process variable data is obtained from means such as online measurements made in the various units of the recovery boiler system, computation of process variables using the process model, use of laboratory data and from combinations of the said means to estimate one or more boiler parameters.

In an embodiment of the present invention, the method for optimization of one or more boiler parameters is applied to parameters comprising of performance parameters such as reduction efficiency, combustion efficiency, steam quality, heat loss; operational parameters such as cost of boiler operation and parameters that are not directly measurable such as sulfate concentration, sulfide concentration in an unit of the recovery boiler system.

In another embodiment of the present invention, the method for optimization of one or more boiler parameters is carried out with formulation of objective function with atleast one boiler parameter comprising of terms describing reduction efficiency, combustibles in flue gas going out of the boiler or the quality of the superheated steam or the excess oxygen control for better heat utilization or a combination of any of these.

As a second aspect of the present invention, a system for optimization of one or more boiler parameters in a recovery boiler process is provided. The system comprises of (i) a process model component having a process model describing relationship between various process variables for one or more units of the recovery boiler system;

(ii) a parameter estimation component to estimate atleast one unit parameter of the recovery boiler system using the said process model component

(iii) an optimization component to perform computation for optimization of one or more boiler parameters using the process model component and the parameter estimation component (iv) a controller component to control one or more boiler parameters having one or more setpoints provided by the optimization component and

Wherein, the process model for atleast one unit of the recovery boiler system is based completely or partially on a first principle mathematical model of the recovery boiler process and the parameter estimation component uses the means such as online measurements made in the various units of the recovery boiler system, computation of process variables using the process model, use of laboratory data and from combinations of the said means to estimate one or more boiler parameters comprising of components that are not directly measurable.

As a third aspect of the present invention, a control system for paper and pulp process is provided. The control system comprises of

(i) a process model component having a process model describing relationship between various process variables for atleast one unit from a first set of units

(ii) a parameter estimation component to estimate atleast one unit parameters of the first set of units using the process model component

(iii) a controller component to control a second set of one or more units based on the estimated atleast one unit parameters and

Wherein, the process model for atleast one unit of the first set of units is based completely or partially on a first principle mathematical model and the parameter estimation component uses the means such as online measurements made in the various units of the recovery boiler system, computation of process variables using the process model, use of laboratory data and from combinations of the said means to estimate one or more boiler parameters comprising of components that are not directly measurable.

As an embodiment of the present invention, the control system in a paper and pulp process has the first set of units that are of recovery boiler system and the second set of units are of green liquor clarifier or causticizer system.

As another embodiment of the system for control in a paper and pulp process describes the first set of units and the second set of units to have atleast one unit in common.

As fourth aspect of the present invention, a system for estimation and prediction of one or more boiler parameters for a recovery boiler is provided. The system comprises of (i) a process model component having a process model describing relationship between various process variables of recovery boiler system

(ii) a parameter estimation component to estimate atleast one boiler parameter using the process model component (iii) a prediction component to obtain one or more trajectories of boiler parameters over a defined prediction time period using the process model component and the parameter estimation component and

Wherein, the process model for atleast one unit of the recovery boiler system is based completely or partially on a first principle mathematical model of the recovery boiler process and the parameter estimation component uses the means such as online measurements made in the various units of the recovery boiler system, computation of process variables using the process model, use of laboratory data and from combinations of the said means to estimate one or more boiler parameters comprising of components that are not directly measurable.

In another embodiment of the present invention, the relationship defined in the process model, parameters estimated by the parameter estimation component or parameters predicted by the prediction component is used to develop a objective function or used as a constraint for obtaining optimization solution.

In another embodiment of the present invention, the methods and systems as described in the invention use process models formulated from means such as that based completely or partially on first principles; empirical rules or data; stochastic; algorithmic models or the combinations as hybrid models.

BRIEF DESCRIPTION OF THE DRAWING

It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.

FIG. 1 shows a typical recovery boiler process and the various units of a recovery boiler process FIG. 2 shows schematic representation of online-optimization system of recovery boiler process; FIG 3 shows schematic of control system for paper and pulp process

DETAILED DESCRIPTION OF THE INVENTION

The invention describes a method and a system that predicts the process variables such as concentration of chemical species involved in the recovery boiler processes with improved accuracy. The method and the system is extended beyond the recovery boiler system to improve control of many other units in the paper and pulp process.

Fig. 1 describes a conventional recovery boiler system 100. The recovery boiler system 100 can be considered comprising of various units, the Furnace 110 (Combustion chamber), Superheater 120, Economizers 130 and Steam Drum 140. The recovery boiler system 100 is also referred to as recovery boiler plant.

(A) Furnace- The unit of the recovery boiler where the black liquor 102 is fired and burnt is the furnace 110. There is a bed of smelt 108 at the bottom of the furnace 110 where the reduction reactions take place that reduces the sulfates to sulfides (Na ₂ SO ₄ + 4C = Na ₂ S + 4CO). The carbonaceous material in black liquor 102 is gasified and oxidized by the combustion air 104 entering the furnace 110 at different locations. Air 104 is also expressed as oxygen and interchangeably used in the description. The component analysis of the black liquor 102 is treated as an important aspect in model formulation. Since, the composition of the black liquor

102 remains almost constant, continuous input for component analysis is not required. The fractions of various components of the black liquor 102 are treated as being deterministic. Black Liquor 102 is considered to have components such as Sodium, Carbon, Hydrogen, Sulfur, Oxygen, Nitrogen, and Water. It is assumed that 50 % of sulfur is released as H ₂ S during devolatilization of black liquor 102 inside the furnace 110. The remaining sulfur is assumed to be present in the form of Na ₂ S and Na ₂ SO ₄ at a known molar ratio. This initial concentration ratio of Na ₂ S and Na ₂ SO ₄ in the black liquor char is set to be 1:1. The remaining sodium is assumed to form Na ₂ Cθ3. During devolatilization of black liquor inside the furnace 110, a known fraction of carbon is assumed to be released as different gaseous compounds. The fraction of carbon in black liquor 102 that goes into these compounds is assumed to be 50%, which is based on small scale experiments. In the current model the carbon containing gases are assumed to be CO, CO ₂ and CH ₄ . The total amount of CO and CO ₂ in the volatile gases depends upon the available oxygen chemically bound in the virgin black liquor still left after formation of the oxygen containing components (Na ₂ CO ₃ , Na ₂ SO ₄ ), as described above. A mole ratio for [CO]/[CO2] is defined as a constant for model formulation. Hydrogen present in black liquor

102 is assumed to be released as H ₂ .

The rate of the reaction for heterogeneous reactions in the furnace 110 is treated to be dependent on both mass transfer and the reaction kinetics. The rate of mass transfer is assumed to be inversely proportional to the square of the diameter of black liquor 102 drops. The sulfate reduction rate is primarily controlled by the reaction kinetics. The sodium sulfide present in the black liquor solids and in the smelt is assumed to be easily oxidized in contact with oxygen.

For the gas phase reactions, it is assumed that the reactions involving CH ₄ and H ₂ are very fast and all of hydrogen and methane is consumed first. The rate of oxidation of CO is assumed directly proportional to the concentration of CO and O ₂ inside the furnace 110. The mathematical model of the furnace 110 is described by equation (eq 1) below dx ., .

— = f(x,p,y) (eq l)

Term x, in equation (eq 1), is the variable of interest such as temperature, pressure and concentrations of various chemical species like Na ₂ SO ₄ , Na ₂ S, NaOH, Na ₂ CO ₃ , CO, CO ₂ inside the furnace 110. Function f described byf(x,p,y) in equation (eq 1) is the mathematical representation of phenomena affecting the variable x in the furnance. Term/? in equation (eq 1) represents a constant associated with structural aspects of the furnace, eg volume of the furnace.

Term y, in equation (eq 1) represents physical factors that affect f(x, p, y) . The physical factors are described in the model as per equation (eq 2) as a function 'g' of term 'x' and 'p'. y = g( ^χ >p) ( ^e q 2)

An example of physical factors is droplet size of the black liquor 102 entering the furnace 110 through nozzles located at the walls of the furnace 110. The black liquor 102 droplet size is calculated in the model based on the viscosity of the black liquor and the nozzle size. The droplet size determines the rate of gasification reactions that convert carbon in the black liquor to CO and CO ₂ and the rate of reduction reactions.

The solution of these equations gives the concentration of various components and variables of interest such as temperature inside the furnace 110.

(B) Superheater - The furnace 110 is followed by a series of heat exchangers to transfer the heat from the hot flue gas 106 to the water. Superheater 120 is the heat exchanger that experiences the hottest flue gas because it is next to the furnace 110. This unit of recovery boiler is used to superheat the steam coming from steam drum 140, and hence called the superheater 120. The mathematical model of superheater is also represented in similar manner as in equations (eq 1) and (eq 2). The model includes the heat balance equations that determine the rate of heat transfer from flue gas 106 to the saturated steam 125 and hence determines the temperature of the superheated steam 127 leaving the superheater 120. Attemperators (not shown the figure) are used to maintain the temperature of the steam by showering water.

(C) Steam Drum 140 is that unit of recovery boiler that receives the radiative heat from the furnace 110 and thus generates the steam from the water 137 coming from the economizers 130. The mathematical model of the steam drum 140 includes the energy balance and the two phase equilibrium equation for steam-water to predict the rate of evaporation. The pressure difference between the steam drum 140 and superheater 120 is calculated which is used to determine the flow rate of the saturated steam 125 out of the steam drum. The mathematical model of superheater is also represented in similar manner as in equations (eq 1) and (eq 2).

(D) Economizer 130 is the unit of recovery boiler that is used to preheat the feedwater 135 before it goes to the steam drum 140. The heat exchange takes place between the flue gas 106 coming from the superheater 120 and the feed water 135. Typically, there are multiple units of economizer to heat the water near to its saturation temperature. The model of the economizer

130 includes the energy balance equations. The heat transfer is considered to occur via convection between the flue gas 106 and feed water 135. The mathematical model of superheater is also represented in similar manner as in equations (eq 1) and (eq 2).

The following terms are used throughout the description, the definitions of which are provided herein to assist in understanding various aspects of the subject innovation. It is to be understood that this definition is not intended to limit the scope of the disclosure and claims appended hereto in any way.

Reduction efficiency refers to the efficiency of the reduction reactions taking place in the boiler that convert sulfates, present in the boiler feed, to sulfides.

The reduction efficiency is calculated with

Sulfide J

Reduction Efficiency = | sulfide + sulfate) smelt

Combustion efficiency refers to the efficiency of the combustion process of the organics inside the furnace section the recovery boiler.

Steam quality refers to the temperature and pressure of the superheated steam formed inside the recovery boiler.

Reduction efficiency is the measure of conversion OfNa ₂ SO ₄ to Na ₂ S. This reduction reaction occurs in the char bed at the bottom of the recovery boiler. Like any other reaction rate, the reduction rate depends on the temperature of the bed. Due to unavailability of online measurements OfNa ₂ S and Na ₂ SO ₄ , it is difficult to predict and control the reduction efficiency.

Current practice of controlling (or maximizing) reduction efficiency uses the bed temperature as the only variable to get an indirect measure of the reduction efficiency. But, reduction efficiency also depends on several other factors such as the availability of oxygen at the bottom of the bed.

Another factors that affect reduction phenomenon are, the chemical composition of the black liquor entering the furnace of recovery boiler, drying, volatilization and combustion reactions that take place during the flight of black liquor from the liquor nozzles to the char bed. All these phenomena mentioned above affect the concentration of the chemical species (carbon, NaOH, Na2S, Na2CO3 and Na2SO4) that reach the char bed and hence affect the reduction reaction rate.

If the stream that is fed back to the digester from recovery processes contains more of undesirable species (Na ₂ CO3 and Na ₂ SO ₄ ), the efficiency of the digester goes down and the mill end up using more chemicals for achieving the same quality of pulp produced from digester. Hence, the reduction efficiency (of reduction reaction, Na ₂ SO ₄ to Na ₂ S) should be maintained as high value as possible ~99 % and the variations in it should be minimized so as to minimize the variations in the quality of the pulp produced from the digester.

The combustion efficiency is monitored via either the level of pollutants such as CO, NO _x in the flue gas leaving the recovery boiler plant 100 or the level of excess oxygen measured in the flue gas leaving the recovery boiler plant 100. The measurements of the pollutants, particulate carbon, excess level of oxygen in the flue gas are some performance parameters that can be used to monitor the combustion efficiency of the recovery boiler plant 100.

These aspects are important in defining the rules for optimization of performance and operation of the recovery boiler system 100.

The recovery boiler process model formulated as described earlier is used for optimal control of performance and operation of the recovery boiler system 100. One of the preferred formulations of the obj ective function is e ^T W _e e + Au ⁷ W _n Au subject to constraints

Where, e = error vector

Δu = vector of the changes in manipulated variables 'u' W _e = weight matrix for error vector

W _u = weight matrix for manipulated variables

The error vector 'e' in the object function includes various terms that quantify the difference in the desired and actual performance of the recovery boiler system. One of the ways of defining the error vector is as follows Actual reduction efficiency - Desired reduction efficiency Actual Steam Temperature - Desired Steam Temperature e = Actual level of combustibles - Desired level of combustibles Actual level of excess oxygen - Desired level of excess oxygen Actual char bed height - Desired char bed height

The above formulation of error vector ensures that the combustibles, steam temperature and reduction efficiency are maintained at the desired level along with the optimum usage of air (oxygen) and optimum char bed height. It is to be noted that this formulation is to optimize performance of the recovery boiler system.

Other formulations include operational parameters such as cost of recovery boiler operation or the formulation for performance is used in combination with operational parameters of recovery boiler operation.

It is to be noted that some of the variables and parameters used in the error vector like steam temperature, level of combustibles are measured and some other variables and parameters like reduction efficiency, char bed height are estimated from the model. It is also to be noted that for some of the variables such as char bed height there is a choice to either compute it using the model incase no direct measurement is made. The variables or parameters that are estimated by use of the model are termed as estimated parameters.

The parameters such as performance parameters, operational parameters, measured and estimated parameters are termed as boiler parameters. Parameters concerned with a specific unit of the recovery boiler system or that of the paper and pulp industry is referred to as unit parameters.

The constraints for the optimization are (i) Furnace Pressure < P

(ii) Black liquor temperature > T

(iii) Qa <Black Liquor flow rate < Qi

(iv) F ₂ <Air flow at various levels < Fi

(v) Black Liquor Droplet size > D (vi) Steam drum level > L

These constrained variables are either measured directly in the recovery boiler system or is inferred from the recovery boiler model. P, T, Ql, Fl, D and L are reference quantities. The manipulated variables (denoted as 'u' in objective function) are chosen to keep the outputs at desired levels or to keep the error (e) at minimum. These manipulated variables are the process variables of the recovery boiler. Some examples of the manipulated variables are Black Liquor Temperature, Air flows, Air Temperature, Inlet Water flow to economizers and Attemperator flow. The term process variable is used to indicate all the variables associated with the process model for which relationships are defined through the definitions in the process model. The value of the process variable is termed as process variable data.

Figure 2 illustrates the system 200 for control and optimization of recovery boiler process. A first principle based process model of the recovery boiler system 100 is used as a central component for control and optimization of the recovery boiler system 100. The process model is available in form of a software module known as process model component 210. Another components known as parameter estimation component 220 is developed for estimation of certain boiler parameters that are either not measurable or needs some form of computation for estimation. The parameter estimation component 220 is developed using the relationships defined in the process model or formulations as defined for performance parameters. The parameter estimation component also uses data from online measurements and laboratory analysis 225 (data entered in the system at periodic intervals) to fine tune process model parameters (also referred as model parameters) of the process model to ensure close match between the actual process/ phenomena in the recovery boiler system. Such a fine tuning is done on a regular basis as found suitable for operation. The online and laboratory measurements is also used by the estimation module 220 to calculate the concentration of the chemical species, Na ₂ SO ₄ , Na ₂ S, NaOH, Na ₂ CO ₃ , in the smelt stream 108 going out from the bottom of the recovery boiler furnace 110. The parameter estimation component 220 also computes boiler performance parameters described in terms of reduction efficiency, combustion efficiency and the steam quality by using the recovery boiler model 210. Such estimations help in achieving better control of performance and operational parameters of the recovery boiler system.

It is to be noted that though the description is made with reference to model formulation based on first principles for the process model component and parameter estimation component, various other kinds of model may be used to achieve some or all the objectives mentioned in this invention. Example of these models are models that are partially based on first principles, models based on empirical rules or data; data driven models, stochastic models; algorithmic models such as those based on neural network, genetic algorithms and the combinations as hybrid models.

The optimization component 230 uses the process model component 210 and parameter estimation component 220 to compute solution for the optimization problem. The solution is expressed as setpoint to one or many regulatory controller 240 that control the recovery boiler processes.

In another aspect of the invention, the control system for the recovery boiler system has a prediction component 320 depicted in Figure 3. The prediction component uses the recovery boiler process model 210 to predict the future trend of boiler parameters or unit parameters in the recovery boiler system. The prediction component 320 takes the current state of the recovery boiler plant 100 as an input from the estimation module 220 and then predicts performance parameters over a finite future time horizon. Thus, a prediction component is used to obtain one or more trajectories of boiler parameters over a defined prediction time period using the process model component and the parameter estimation component. The predicted values may be monitored by the production manager to assess the current and future performance of the recovery boiler plant 100. The prediction information may also be used by the optimization component 230 to solve the optimization problem.

The prediction component 320 output is also available for controllers controlling the operations of units outside the recovery boiler system for example the downstream units 310. The predicted values from the prediction component are used to provide set-points to the downstream units such as green liquor clarifier and causticizer unit. The knowledge of the composition of the smelt 108 coming out of the recovery boiler in particular will help in stabilizing and controlling the alkali content of the liquor leaving the smelt dissolving tank which in turn results in improved performance of the causticizer unit. Since the residence times in the smelt dissolving tank, and clarifier tanks are large, on-line indication of the smelt 108 compositions is of great help in stabilizing and controlling the composition of the green liquor going to the causticizer unit from clarifier unit, both downstream of the recovery boiler 100. It is to be noted that the green liquor clarifier unit and the causticizer unit may comprise of multiple units, thus forming green liquor clarifier system and causticizer system respectively. Information such as smelt 108 composition from the performance monitoring system 320 is used as feed forward signals to control of the density/alkali of green liquor after the clarifier. Thus, information from units of the recovery boiler 100 is used to control and optimize units of the paper and pulp mill that are outside the recovery boiler system 100. This information from units of the recovery boiler 100 is also useful to other systems that have at least one unit in common with the recovery boiler system.

Thus, the developed method and a system helps to accurately estimate and predict the process variables including the concentration of chemical species in the recovery boiler system and thereby efficiently control and optimize the recovery boiler processes and other associated processes. Various other modifications and alterations in the structure and method of operation of this invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments and specifically for the recovery boiler plant and downstream units used in paper and pulp industry, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments or a particular plant system used in paper and pulp industry. It is intended that the following claims define the scope of the present invention and that structures and methods within the scope of these claims and their equivalents be covered thereby.