Method and apparatus for the energy-efficient operation of secondary dust removal systems

Abstract

The invention relates to a control method for a secondary dust removal system in which a pipe network connects an induced draft fan to at least two suction points. The pipe network comprises a controllable exhaust air flap for each suction point, the position of said flap influencing the volumetric flow rate at the suction point. A mathematical model of the pipe network allows the method to energy-efficiently control the exhaust air flaps and the induced draft fan.

Claims

1. Control method for a secondary dust removal system (400) in which a pipeline network connects an induced draft fan to at least two suction points, wherein the pipeline network comprises a controllable exhaust air flap for each suction point, whose flap position influences a volumetric flow rate at the suction point, and wherein the control method is characterized in that it involves the following steps: providing a mathematical system model (300) which describes the pipeline network based on physical and geometrical properties (10) of all of the pipeline network elements and their functions; calculating pressure losses in the pipeline network on the basis of the fully determined mathematical system model and the volumetric flow rates required at the suction points (20); calculating of the flap position for each exhaust air flap and a pressure loss for the secondary dust removal system which allows the required operating volumetric flow rates to be achieved with the least possible delivery power (30); controlling of the exhaust air flaps so that the calculated flap positions are assured (40); controlling of the induced draft fan so that its rotary speed is increased until the calculated pressure loss for the entire system is achieved and the volumetric flow rates are assured at the suction points (50), wherein the physical properties of the pipe line elements are based on following quantities: coefficients of resistance characterizing the pipe elements through which the flow occurs; geometrical properties of the pipe elements through which the flow occurs; coefficient of heat transfer of the pipe elements through which the flow occurs; volumetric flow rates through the pipe elements.

2. Control method according to claim 1, wherein the pipeline network comprises a main string, which is connected to the induced draft fan and which branches into at least two secondary strings, each of which is connected to at least one suction point.

3. Control method according to claim 2, wherein the calculation steps (20, 30) include the following additional steps: calculating of the pressure losses in the main and secondary strings based on the given geometrical and physical properties and the volumetric flow rates calculating of the pressure loss of the entire system by the mesh rule.

4. Control method according to claim 1, wherein the pipeline network is composed of various pipe elements which include straight pipe elements, curves, expansions, reductions, and merging points.

5. Control method according to claim 1, wherein the step of controlling of the induced draft fan includes increasing its rotary speed from the position of rest.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) In what follows, the figures of the sample embodiments are briefly described. Further details will be found in the detailed description of the sample embodiments. There are shown:

(2) FIG. 1, a process flow chart of the method according to the invention;

(3) FIG. 2, a schematic representation of a preferred embodiment of the device according to the invention;

(4) FIG. 3, a schematic representation of a pipe element;

(5) FIG. 4, the coefficient of resistance determined for the pipe element of FIG. 4, dependent on the volumetric flow rate through the element;

(6) FIG. 5, the coefficient of resistance determined for a controllable exhaust air flap, dependent on the closing position of the flap;

(7) FIGS. 6a and 6b, a node in a pipeline network and the equivalent resistance network;

(8) FIGS. 7a and 7b, a mesh in a pipeline network and the equivalent resistance network;

(9) FIG. 8a, a diagram of a sample pipeline network of a secondary dust removal system;

(10) FIG. 8b, a diagram of a resistance network which characterizes the pipeline network of a secondary dust removal system.

DETAILED DESCRIPTION OF THE SAMPLE EMBODIMENTS

(11) The new control concept according to the invention is based on the idea of forming a mathematical model of the secondary dust removal system and performing a computation with the help of provided plant data, resulting in the flap positions and the overall pressure loss of the system. On the basis of these computations, the exhaust air flaps of the system are moved into the corresponding position and the rotary speed of the induced draft fan is increased from an original position of rest until the pressure loss of the overall system produced in this way (measured in the main channel) reaches the level of the calculated overall pressure loss.

(12) In a first step, a mathematical system description is represented on the basis of a graph theory approach and the plant data in a control system software. This corresponds to step 10 in FIG. 1.

(13) In the next step, an algorithm factoring in the system properties and process requirements computes the pressure losses in the system (step 20) and, on this basis, the optimal position of each flap and the overall pressure loss of the system (step 30). On the basis of this provided information, the flaps of the system are placed in the respective position (step 40) and the rotary speed of the induced draft fan is increased until the pressure loss computed by the algorithm is achieved (step 50).

(14) The result is a control system which makes it possible to adjust the previously defined volumetric flow rate at each suction point and reduce to a minimum the overall pressure loss in the system.

(15) This describes the fundamental principle of the invention. The further explanations describe preferred embodiments of the invention and allow the skilled person to implement the method.

(16) As is shown in FIG. 2, the invention likewise involves a device 100, which serves to implement the method. The device comprises a storage element 110, a computer device 120, and a control element 130. In the storage element, a mathematical model 300 of the pipeline network of a secondary dust removal system is saved. With the aid of this data, a computer device 120 calculates the control variables in harmony with the method according to the invention. The resulting flap positions are relayed via the control element 130 to the secondary dust removal system 400. The rotary speed of the induced draft fan of the dust removal system 400 is increased until the pressure loss of the overall system as detected by a suitable sensor agrees with the calculated pressure loss of the computer device 120. The control element 130 increases or decreases the rotary speed for the respective process stage.

(17) In a preferred embodiment, the mathematical model is a graph theory representation of a secondary dust removal system, which can be described and computed by the memorized physical quantities and functions, as well as geometrical dimensions of the various pipeline elements. The resulting model can be described as a resistance network.

(18) Fundamentals of the Mathematical System Design

(19) In order to translate pipeline networks into a mathematical model, it is important to identify and calculate significant physical quantities.

(20) It should be noted that three simplifications are assumed in this description for an efficient computation. These simplifications allow the reader to follow the mode of functioning of the method, yet without limiting the latter in any way. It is assumed that the gas composition for dry air (78.08% N.sub.2; 20.95% O.sub.2; 0.93% Ar; 0.04% CO.sub.2) prevails at each suction point. A specific gas constant of R.sub.S=287.058 J/(kgK) is assumed. The absolute pressure is assumed to be constant (p.sub.Abs=101325 Pa) for the calculation of the density at various places in the network. The pipeline network undergoes no temperature loss, i.e., an adiabatic system behavior is assumed.
Behavior of the Physical Quantities in the System

(21) Determination of the Density of the Suctioned Gas

(22) The density of the suctioned gas plays a major role in the calculation of pipeline networks, since volumetric flow rates also vary with increasing or decreasing density. The density of the suctioned gas can be calculated by the following equation 2

(23) $\begin{array}{cc}=\frac{p_{\mathrm{Abs}}}{R_s{T}_{\mathrm{Abs}}}& \left(2\right)\end{array}$

(24) As is evident from the equation, the absolute ambient pressure p.sub.Abs the ambient temperature T.sub.Abs, and the specific gas constant R.sub.S have influence on the density of the gas. The specific gas constant, as well as the absolute pressure, can be seen as constant quantities in the mentioned method.

(25) Pressure Loss Due to Flow Resistances

(26) When a medium flows through a physical body, the latter presents a resistance, which results in a pressure loss. For the controlling of the flap positions, the pressure losses of the different pipeline strings are detected and included in the mathematical system description. For a unified calculation, the pressure loss of the different components of a dust removal system is used as a nondimensional factor with the help of the coefficient of resistance zeta in the formulas). This factor is a measure of how much of a pressure loss is caused by a component when a flow occurs through it. The following equation 3 shows how the pressure loss is dependent on the coefficient of resistance, the gas density and the flow velocity of the medium.

(27) $\begin{array}{cc}p=\frac{1}{2}{v}^2& \left(3\right)\end{array}$

(28) How the zeta values of different components are determined will be described in the following description.

(29) Converting of Normal Volumetric Flow Rate into Operating Volumetric Flow Rate

(30) It is customary to use so-called normal volumetric flow rates to indicate the required volumetric flow rates in dust removal systems. These refer to a theoretical and idealized comparison value. Usually the following standardized values are used for normal volumetric flow rates [6]. normal temperature=273.15 K normal pressure=101325 Pa

(31) Since secondary dust removal systems are generally not operated at this point, one must compute the actual operating volumetric flow rate for a deviating temperature. With the assumed simplification that constant ambient pressure prevails at every point, the operating volumetric flow rate {dot over (V)}.sub.B can be computed from the normal volumetric flow rate {dot over (V)}.sub.N and the ratio of the operating temperature T.sub.B to the normal temperature T.sub.N (equation 4).

(32) $\begin{array}{cc}{\stackrel{.}{V}}_B={\stackrel{.}{V}}_N\frac{T_B}{T_N}& \left(4\right)\end{array}$

(33) Temperature Change Due to Mixing of Volumetric Flow Rates

(34) If volumetric flow mingling occurs in a secondary dust removal system, with several volumetric flow rates of different temperatures, the new mixture temperature should be determined for further computations. With the described simplification that the same gas composition prevails at each suction point, one gets equation 5, which determines the mixture temperature by a simple ratio calculation.

(35) For this, the individual normal volumetric flows are weighted with the respective temperature and divided by the sum of all normal volumetric flows:

(36) $\begin{array}{cc}{T}_{\mathrm{neu}}=\frac{{\stackrel{.}{V}}_{N1}{T}_1+\mathrm{.Math.}+{\stackrel{.}{V}}_{\mathrm{Nn}}{T}_n}{{\stackrel{.}{V}}_{N1}+\mathrm{.Math.}+{\stackrel{.}{V}}_{\mathrm{Nn}}}& \left(5\right)\end{array}$
Determination of the Required Coefficients of Resistance

(37) In this section we shall discuss the calculation of coefficients of resistance of various pipeline elements. The coefficient of resistance plays a significant role in the calculation of pipeline networks, since is it a direct measure of the causative pressure loss of a pipeline element and also has influence on the dimensioning of pipeline elements.

(38) The following equation 6 shows the definition of the coefficient of resistance.

(39) $\begin{array}{cc}=\frac{2p}{v^2}& \left(6\right)\end{array}$

(40) In order for the determination of the coefficient of resistance to be useful in practice, it is recommended to determine this value empirically, since a calculation of this value would be very complicated on account of the influence of many physical quantities. As follows from equation 6, for an empirical determination of the coefficient of resistance one must measure the pressure loss and the flow velocity. The density of the medium can be calculated by equation 2.

(41) Practical Determination of the Coefficients of Resistance of Partial Sections

(42) For sections of pipeline elements which experience no direct volumetric flow mingling or volumetric flow separation the coefficient of resistance can be assumed to be a constant quantity. In order to keep the measurement and computation expense as low as possible, the longest possible partial sections with different pipeline elements are combined into a single coefficient of resistance. A partial string, for example, can consist of several pipeline elements such as straight pipe elements, curves, expansions and T-pieces.

(43) By the measurement of the absolute pressures at the inlet p.sub.1 and outlet p.sub.2 of the partial string, one can determine the pressure drop in the partial string with the help of equation 7.
p=p.sub.1p.sub.2(7)

(44) By means of a Pitot tube, furthermore, the flow velocity in a partial string can be measured. If the measured pressure drop, as well as the determined flow velocity, is inserted into equation 6, one gets the coefficient of resistance of the partial section.

(45) Practical Determination of the Coefficients of Resistance of T-Pieces

(46) T-pieces have dynamically varying coefficients of resistance. At a volume flow mingling or volume flow separation, a nonlinear relation exists with the resulting pressure loss. FIG. 3 shows a volume flow mingling in a T-piece.

(47) In the technical literature on fluidics it is customary to relate the coefficient of resistance of a T-piece to the overall flow velocity of the mingling partial flows. This procedure is not usable for the control concept presented in this work, because the coefficients of resistance must be coordinated with the respective partial strings in order to compute the flap positions. It is recommended to likewise determine empirically the function of the coefficient of resistance, since the available functions in the literature are subject to large constraints. Due to these constraints, the coefficients of resistance so determined differ greatly from the true values.

(48) In the empirical determination, one forms the ratio of the partial volume flow and the total volume flow. For this condition, a pressure loss measurement per equation 7 is carried out and the flow velocity in the partial string is measured.

(49) If one performs this measurement for different volume flow ratios and computes for each ratio the coefficient of resistance with equation 6, one gets for example the result in FIG. 4. This figure shows the coefficient of resistance of the straight section of the T-piece, plotted against the volume flow ratio.

(50) For an automated control system, this characteristic curve must be translated into a function, so that a verdict as to the magnitude of the coefficient of resistance is obtained for each volume flow ratio. For this, the curve is approximated by a function. This conversion is done with the aid of algorithms which are known in themselves and requires no further description in this regard. For the example in FIG. 4, one gets the following equation 8.
(x)=0.2764x.sup.2.74(8)

(51) The specific behavior of a pipe element depends, of course, on the components used and can be determined by means of the described steps for any given pipe components and especially for any given merging element.

(52) Practical Determination of the Coefficients of Resistance of Exhaust Air Flaps

(53) The coefficient of resistance of exhaust air flaps is primarily dependent on the closing position of the flaps. It is therefore advisable to perform a measurement of the coefficient of resistance in dependence on the closing position. The empirical determination of the resistance is necessary, since the coefficients of resistance in the literature and in various simulation programs differ greatly from one another. For the determination of these values, a characteristic curve is likewise used by the same method as was described for the T-piece.

(54) In order to determine the coefficient of resistance of a flap, the flap is moved in succession to various closing positions. For each closing position, the resulting pressure loss as well as the flow velocity are measured. Then, with equation 6, the coefficient of resistance can be calculated. The following FIG. 5 shows a sample plotted curve of a single-vane exhaust air flap.

(55) If the curve of the example in FIG. 5 is approximated as a function, one gets equation 9.
(x)=0.1061e.sup.0.0776x(9)

(56) Since the control algorithm of the invention requires the function of the closing position in dependence on the coefficient of resistance, the inverse function (equation 10) is formed from equation 9.
(x)=12.8866(ln(x)+2.24337(10)

(57) This function is suitable to being saved in the control algorithm With it, the closing position of the flap can be determined for a required coefficient of resistance.

(58) Transformation of a Pipeline Network into a Mathematical System Model

(59) Derivation of the Quadratic Resistance Law

(60) By analogy with electrical engineering, it is possible to convert pipeline networks into resistance networks. For this representation, it is necessary to transform the various elements of a pipeline network into resistances. This is done by the pressure loss equation for turbulent flows after Darcy (equation 11). This describes the pressure loss in straight pipe sections. Here, represents the coefficient of friction of the pipe and d the diameter of the pipeline element.

(61) $\begin{array}{cc}p=\frac{l}{2d{A}^2}{\stackrel{.}{V}}^2& \left(11\right)\end{array}$

(62) If the pressure loss computation is done with the aid of the coefficient of resistance, equation 11 can be converted into equation 12.

(63) $\begin{array}{cc}p=\frac{p}{2A^2}{\stackrel{.}{V}}^2\mathrm{with}=\frac{l}{d}& \left(12\right)\end{array}$

(64) From this equation, the quadratic resistance law (equation 13) can be derived, since the pressure loss p is equal to the product of the resistance R and the square of the volumetric flow rate {dot over (V)}.

(65) $\begin{array}{cc}p=R{\stackrel{.}{V}}^2\mathrm{with}R=\frac{}{2A^2}& \left(13\right)\end{array}$

(66) With the help of equation 13, it is possible to transform the various pipeline elements into a resistance value.

(67) Node and Mesh Rules in Pipeline Systems

(68) By analogy with the Kirchhoff rules of electrical engineering, similar rules apply in a pipeline network. These rules are used in the control algorithm in order to compute the network.

(69) Node Rule in Pipeline Systems

(70) If a node point occurs in a pipeline network, as in FIG. 6a, volume flows will merge or separate there. The corresponding resistance network is shown in FIG. 6b.

(71) The node rule states that the addition of all inflows and outflows in a node results in a value of zero. Volume flows coming into the node are given a positive value. Outgoing volume flows are given a negative value. Mathematically, the following summation formula (equation 14) expresses this relationship, where k is the number of adjoining partial strings.

(72) $\begin{array}{cc}\underset{n=1}{\overset{k}{.Math.}}{\stackrel{.}{V}}_n=0& \left(14\right)\end{array}$

(73) With the help of this node rule, it is possible for known volume flow inputs to calculate the size of the volumetric flow rate of each partial string.

(74) Mesh Rule in Pipeline Systems

(75) The mesh rule for pipeline systems states that the sum of all pressure losses of a mesh results in a value of zero. FIG. 7a shows such a mesh. FIG. 7b shows the corresponding resistance network. As can be seen in FIG. 7b, for this one determine a mesh in the network and travels in clockwise direction. Pressure losses in the clockwise direction have a positive value, while pressure losses in counterclockwise direction have a negative value.

(76) This relationship is expressed in a formula by equation 15. Here, 1 indicates the number of branches belonging to a mesh.

(77) 0$\begin{array}{cc}\underset{n=1}{\overset{l}{.Math.}}{p}_n=0& \left(15\right)\end{array}$

(78) With the help of the mesh rule, it is possible to calculate hitherto unknown pressures in a pipeline system.

(79) Dividing Up the Partial Strings into Resistances

(80) If one considers a partial string of a secondary dust removal system, it will be seen that it consists of a plurality of different pipeline elements such as curves, expansions or reductions.

(81) Each partial string is divided up into various partial resistances, distinguishing the following. Dynamic resistances of exhaust air flaps, here the coefficient of resistance is a function of the closing position (%); Dynamic resistances of T-pieces, here the coefficient of resistance is a function of the volume flows; Constant resistances of other pipeline elements.

(82) On the basis of these rules, it is obvious that a pipeline network can be converted into a mathematical system model, e.g., a resistance network, which describes the pipeline network. In the control concept according to the invention, the pressure losses of the various partial strings are computed by using the quadratic resistance law and the mesh and node rules. The described rules make it possible to describe pipeline networks which connect an induced draft fan to a plurality of suction points, wherein a main string connected to the induced draft fan can branch multiple times like a tree, and wherein at the end of each branch a suction point is connected. By iterative application of the node and mesh rules, the pressure relations in the overall tree structure are calculated.

Calculation Example

(83) The calculation shown is explained by an imaginary pipeline network with four suction points 1-4 and one induced draft fan 6. For example, a T-piece 5 is represented as a merging piece. This computation is such as is carried out preferably by an arithmetic unit of a computer or a control system. FIG. 8a shows a schematic representation of a secondary dust removal system. On the basis of a graph theory approach, such a network can be converted into an edge and node relation (FIG. 8b). The foundation for a mathematical system description is created by weighting the edges K1-K7 with the physical system quantities and functions, as well as by known geometrical dimensions of the individual components. Such system quantities can be, for example, the following listed properties: geometrical dimensions of the pipeline elements; fixed coefficients of resistance of the pipeline elements; dynamic coefficient of resistance functions of the pipeline elements; heat transfer coefficient of the pipeline elements; pressure increase or pressure decrease due to leakage, booster blower, or the like.

(84) If this information is saved in the storage unit of the corresponding computer or control unit and a complete description of the nodes and edge relations as well as their affiliation is in hand (e.g., in the form of an adjacency matrix), the computer unit can start to determine the following values for the nodes N1-N8 in the system with the help of an algorithm and the specification of the required normal volume flows and temperatures at the suction points: temperature in the pipeline elements; density in the pipeline elements; operating volume flows.

(85) The calculation is thus done by known physical formulas or formulas adapted to this instance (such as Riechmann mixing rules). As the outcome of the computation, one gets the operating volume flow, the density of the suctioned medium, and the temperature of the medium at each node point.

(86) In order to operate such a secondary dust removal system with energy efficiency, the suction point which causes the highest pressure loss must be known. This is ascertained with the help of the information from the weighted edges and nodes. A determination is made as to which path from the induced draft fan to a suction point causes the highest pressure loss (with flap fully opened). For this, the resistance values of the various pipelines are determined by the quadratic resistance law for turbulent flows. These resistance values are multiplied by the square of the volume flows of the respective pipe strings. As the result, one gets the pressure loss of the particular pipe string. Since consecutive pressure losses can be added (validity of mesh and node rules in pipeline networks), the highest pressure loss can thus be determined. The exhaust air flap of this string is 100% opened, all other flaps are adapted in their opening position so that the mesh rule continues to be valid for the required normal volume flows.

(87) The above indicated sample embodiments serve primarily for a better understanding and should not be construed as a limitation. The scope of protection of the present patent application results from the patent claims.

(88) The features of the described sample embodiments can be combined with or exchanged for each other. Moreover, the described features can be adapted by the skilled person to existing circumstances or present requirements.