Method for evaluating fouling of a heat exchanger

Abstract

The invention relates to a method for evaluation of fouling of passages of a spacer plate (10) of a tubular heat exchanger (11), wherein first, second and third pressure sensors (31, 32, 33) are arranged, the method comprising steps of: (a) during a transient operation phase of the heat exchanger determination of a value over time of Wide Range Level NGL, from the measurements of the first and third pressure sensors (31, 33), and of a value over time of Narrow Range Level NGE, from the measurements of the second and third pressure sensors (31, 33); (b) determination of a value over time of Steam Range Level deviation ΔNGV, corresponding to the NGL from which a component representative of a variation of free water surface in the heat exchanger has been filtered, from the values of NGL and NGE; (c) comparison of the determined value of ΔNGV with a set of reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, each reference profile ΔNGV.sub.i being associated with a level of fouling so as to identify a target reference profile ΔNGV.sub.opt among the reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, which is that closest to the determined value ΔNGV. (d) restored on an interface (3) of the level of fouling associated with the identified target reference profile ΔNGV.sub.opt.

Claims

1. A method for evaluation of fouling of passages of a spacer plate of a tubular heat exchanger, said passages being arranged along the tubes for fluid to pass through the spacer plate, wherein a first pressure sensor is arranged at a low altitude of the heat exchanger; a second pressure sensor is arranged at a medium altitude of the heat exchanger; a third pressure sensor is arranged at a high altitude of the heat exchanger; the method comprising performing by a data-processing unit steps of: (a) during a transient operation phase of the heat exchanger, determining a value over time of Wide Range Level, NGL, from the measurements of the first and third pressure sensors, and of a value over time of Narrow Range Level, NGE, from the measurements of the second and third pressure sensors; (b) determining a value over time of Steam Range Level deviation, ΔNGV, the Steam Range Level deviation corresponding to the Wide Range Level from which a component representative of a variation in free water surface in the heat exchanger has been filtered, from the values of NGL and NGE; (c) comparing the value of ΔNGV determined with a set of reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, each reference profile ΔNGV.sub.i being associated with a level of fouling, so as to identify a target reference profile ΔNGV.sub.opt among the reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, which is that closest to the determined value ΔNGV; and (d) outputting on an interface the level of fouling associated with the identified target reference profile ΔNGV.sub.opt.

2. The method according to claim 1, wherein the heat exchanger is a steam generator having a heat zone and a steam zone located at an altitude greater than the heat zone, the tubes extending in the heat zone only.

3. The method according to claim 2, wherein the first pressure sensor is arranged substantially at the bottom of the heat zone, the second sensor is arranged substantially at the top of the heat zone, and the third exchanger is arranged substantially at the top of the steam zone.

4. The method according to claim 1, wherein the value over time of ΔNGV is determined from the values of NGL and NGE by the formula ΔNGV=NGL−NGE.

5. The method according to claim 1, wherein the level of fouling is a rate of fouling expressed between 0 and 1.

6. The method according to claim 1, comprising a previous step (a0) for generation of said set of reference profiles ΔNGV.sub.i during said transient operation phase of the heat exchanger.

7. The method according to claim 6, wherein step (a0) comprises performing steps (a) and (b) for a reference heat exchanger similar to said heat exchanger during at least two occurrences of said transient operation phase respectively associated with a first level of known fouling and a second level of known fouling greater than the first level of fouling so as to obtain a first reference profile ΔNGV.sub.Level.sub.low during said transient operation phase of the heat exchanger for the first level of fouling and a second reference profile ΔNGV.sub.Level.sub.high during said transient operation phase of the heat exchanger for the first level of fouling, the other reference profiles ΔNGV.sub.i, during said transient operation phase of the heat exchanger being calculated from the first and second profiles ΔNGV.sub.Level.sub.low and ΔNGV.sub.Level.sub.low.

8. The method according to claim 7, wherein the other reference profiles ΔNGV.sub.i are calculated from the first and second reference profiles ΔNGV.sub.Rate.sub.low=ΔNGV.sub.Level.sub.low and ΔNGV.sub.Rate.sub.high=ΔNGV.sub.Level.sub.high by using the formula $\Delta {\mathrm{NGV}}_i=\frac{\begin{array}{c}\left({\mathrm{Rate}}_{\mathrm{high}}-{\mathrm{Rate}}_i\right).Math.\Delta {\mathrm{NGV}}_{{\mathrm{rate}}_{\mathrm{low}}}+\\ \left({\mathrm{Rate}}_i-{\mathrm{Rate}}_{\mathrm{low}}\right).Math.\Delta {\mathrm{NGV}}_{{\mathrm{Rate}}_{\mathrm{high}}}\end{array}}{\left({\mathrm{Rate}}_{\mathrm{high}}-{\mathrm{Rate}}_{\mathrm{low}}\right)},$ Rate.sub.high, Rate.sub.i and Rate.sub.low being the levels of fouling respectively associated with ΔNGV.sub.Level.sub.high, ΔNGV.sub.i, and ΔNGV.sub.Level.sub.low.

9. The method according to claim 8, wherein, for each level of fouling considered, the step (a0) comprises performing the steps (a) and (b) for said reference heat exchanger similar to said heat exchanger during at least three occurrences of said transient operation phase associated with said level of fouling so as to obtain at least three real profiles ΔNGV.sub.r during said transient operation phase of the heat exchanger for said level of fouling, the obtaining of the reference profile ΔNGV.sub.i, during said transient operation phase of the heat exchanger for said level of fouling comprising calculating an average of the real profiles ΔNGV.sub.r then approximation of said average by a given function.

10. The method according to claim 9, wherein the average of the real profiles ΔNGV.sub.r is approximated by a three-degree polynomial.

11. A non-transitory computer program product on which instruction program code are stored, the instruction program code, when executed by one or more computing device perform a method, the method comprising: (a) during a transient operation phase of the heat exchanger, determining a value over time of Wide Range Level, NGL, from the measurements of the first and third pressure sensors, and of a value over time of Narrow Range Level, NGE, from the measurements of the second and third pressure sensors; (b) determining a value over time of Steam Range Level deviation, ΔNGV, the Steam Range Level deviation corresponding to the Wide Range Level from which a component representative of a variation in free water surface in the heat exchanger has been filtered, from the values of NGL and NGE; (c) comparing the value of ΔNGV determined with a set of reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, each reference profile ΔNGV.sub.i being associated with a level of fouling, so as to identify a target reference profile ΔNGV.sub.opt among the reference profiles ΔNGV.sub.i, for said transient operation phase of the heat exchanger, which is that closest to the determined value ΔNGV; and (d) outputting on an interface the level of fouling associated with the identified target reference profile ΔNGV.sub.opt.

12. A system for evaluating fouling of passages of a spacer plate of a tubular heat exchanger, the system comprising: a data processing unit in communication with a first pressure sensor, a second pressure sensor, and a pressure sensor, the data processing unit configured to: (a) during a transient operation phase of the heat exchanger, determine a value over time of Wide Range Level, NGL, from the measurements of the first and third pressure sensors, and of a value over time of Narrow Range Level, NGE, from the measurements of the second and third pressure sensors; (b) determine a value over time of Steam Range Level deviation, ΔNGV, the Steam Range Level deviation corresponding to the Wide Range Level from which a component representative of a variation in free water surface in the heat exchanger has been filtered, from the values of NGL and NGE; (c) compare the value of ΔNGV determined with a set of reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, each reference profile ΔNGV.sub.i being associated with a level of fouling, so as to identify a target reference profile ΔNGV.sub.opt among the reference profiles ΔNGV.sub.i, for said transient operation phase of the heat exchanger, which is that closest to the determined value ΔNGV; and (d) output on an interface of the level of fouling associated with the identified target reference profile ΔNGV.sub.opt.

13. The system according to claim 12, wherein the first pressure sensor is arranged at a low altitude of the heat exchanger; the second pressure sensor is arranged at a medium altitude of the heat exchanger; and the third pressure sensor is arranged at a high altitude of the heat exchanger.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Other characteristics, aims and advantages of the invention will emerge from the following description which is purely illustrative and non-limiting and which must be considered with respect to the appended drawings, in which:

(2) FIGS. 1a-1b, already commented on, according to two views schematically illustrate a heat exchanger of steam generator type;

(3) FIG. 2, already commented on, in a plan view, illustrates a branched passage in a spacer plate, in which a tube passes, according to a current configuration of a steam generator;

(4) FIG. 3, already commented on, schematically illustrates the instrumentation of the steam generator of FIGS. 1a-1b;

(5) FIG. 4 illustrates architecture for executing the present method,

(6) FIG. 5 is a block diagram of the method for evaluation of fouling,

(7) FIGS. 6a and 6b illustrate examples respectively of signal ΔNGV and NGL;

(8) FIGS. 7a to 7c illustrate the rate of signals used to obtain a set of reference profiles ΔNGV.sub.i.

DETAILED DESCRIPTION OF THE DRAWINGS

(9) Architecture

(10) In reference to FIG. 4, a method is proposed for evaluation of fouling of passages of spacer plates 10 of a tubular heat exchanger 11, said passages 12a, 12b being arranged along the tubes 11 for fluid to pass through the spacer plate 10.

(11) The tubular heat exchanger is preferably a steam generator of the type described in the introduction.

(12) This method is a refinement of dynamic NGL techniques and utilises pressure measurements coming from the first, second and third pressure sensors 31, 32, 33.

(13) In general: The first sensor 31 is at a low altitude of the heat exchanger; The second sensor 32 is at a medium altitude of the heat exchanger; The third sensor 33 is at a high altitude of the heat exchanger.

(14) Given a steam generator having a heat zone 20 in which the tubes 11 and a steam zone 25 extend above the heat zone, then preferably (“high” and “low” extend in terms of altitude): The first sensor 31 is arranged substantially at the bottom of the heat zone 20 (low part of the downcomer 22); The second sensor 32 is arranged substantially at the top of the heat zone 20 (i.e. substantially at the bottom of the steam zone 20); The third exchanger 33 is arranged substantially at the top of the steam zone 25.

(15) The present method is executed by a processing unit 1 (for example one or more processors) of equipment, for example a server, connected to the pressure sensors 31, 32, 33 so as to have measurements of associated pressures.

(16) The equipment can comprise a memory 2 (for example a hard drive) for storing data, and an interface 3 for interaction with an operator.

(17) Delta NGV

(18) The equation of the pressure at the bottom of the steam generator can be expressed as follows:
P.sub.bas=ρdowncomer.Math.g.Math.z.sub.surface+P.sub.dome−pdc
Pdc: loss of reference load in the downcomer
P.sub.base: pressure measured at the level of the first sensor 31
P.sub.dome: pressure measured at the level of the third sensor 33
ρ.sub.downcomer: volume reference mass
z.sub.surface: designates the altitude of the surface of the water (i.e. the water level) relative to the altitude of the first sensor 31.

(19) As explained, fouling increases the resistance of passage of water in the riser 20 as it reduces the passage cross-section of the water, decreasing the steam flow and consequently the flow in the recirculation loop and therefore losses in load Pdc in the downcomer 22 and lowering the value of P.sub.dome.

(20) Variations in the exchanged power of the steam generator, the phase changes inside the heat zone 20 (liquefaction in cooling or evaporation in a power increase) generate violent thermohydraulic transients. These transients will act strongly on the pressure at the base of the steam zone 25 (i.e. at the level of the second pressure sensor 32), sensitive to the downcomer flow rate, as well as on the pressure at the apex of the steam zone 23 (i.e. at the level of the third pressure sensor 33) and as a consequence are particularly adapted to the qualification of losses of internal loads of the steam generator, directly connected to fouling.

(21) However, the pressure measurement at the base of the heat zone 20 (i.e. at the level of the first pressure sensor 31) is also representative of the altimetry of the free surface of the water in the generator. This water height is not representative of the behaviour of the steam generator but solely of the quality of water regulation. As a consequence its variations substantially complicate interpretation of the low-pressure signal of the first pressure sensor 31.

(22) The NGE is conventionally used for fine regulation of the water height (over a range of 1 m) in the steam generator as an operation, and it controls the drinking water inlet valves of the steam generator as a nominal operation of the installation. As is familiar for the skilled person NGE is a function of the difference between the pressure measurements of the second and third sensors 33.

(23) The NGL per se monitors (over a range of more than 15 m) the water level of the steam generator when the plant starts up, when stopped and in abnormal operating situations (when the second pressure sensor 32 is dewatered).

(24) The present method uses a novel indicator called ΔNGV (delta NGV, “steam range level deviation”, “écart de niveau gamme vapeur” in french) which uses the respective sensibilities of the NGE (which provides a clear indicator of the real water level of the steam generator) and of the NGL (sensitive to variations in level, to variations in flow and temperature via the volume mass of the water of the downcomer) to construct an indicator specifically targeting the magnitudes impacted by fouling.

(25) The ΔNGV is based on the NGL from which the component has been filtered due to the variation in free surface (incarnated by the NGE measurement), and it then corrects the NGL, insufficient up to this time to the effective characterisation of fouling. In practice this variation is not in fact correlated to fouling of the steam generator but only to the conduct of the ARE regulation.

(26) This indicator is determined from the values of NGL and NGE preferably by the formula ΔNGV=NGL−NGE.

(27) The values ΔNGV, NGL and NGE are expressed advantageously in mCE (water column metres), or in a pressure unit (bar, pascal, etc.) or as a percentage relative to maximum levels.

(28) As is clear from FIGS. 6a and 6b, the “bump” observable at two thirds of the NGL signal, which is symptomatic of a variation of the real level in the GV, is completely filtered on the ΔNGV signal.

(29) Process

(30) In reference to FIG. 5, the method begins as explained by a step (a) for determination by the data-processing unit 1, during a transient operation phase of the heat exchanger, of a value over time of the Wide Range Level NGL, from the measurements of the first and third pressure sensors 31, 33, and of a value over time of the Narrow Range Level NGE, from the measurements of the second and third pressure sensors 31, 33.

(31) In general, “transient operation phase” means a dynamic phase during which the level of heat energy brought to the heat exchanger fluctuates. A transient operation phase can be defined by an associated value of Active Electric Power (AEP) so as to grasp its kinetics.

(32) Said transient phase is in particular that occurring during regulated standardised periodic assays such as “EP RGL 4” on 900 MW nuclear plants (CP1 and CP2 units) and 1300 MW plants, but many other periodic assays of the EP RGL, EP RCP, EP RPN, etc. family.

(33) It is also quite possible to apply this method during procedures for islanding, automatic stopping of the reactor, turbine triggering, etc. and even as normal operation during load-monitoring transients (power tapping or drop according to the request of the electric network manager).

(34) As will become evident later from the text, the present method uses reference profiles for given transient operating phases (“abacuses”) and it suffices that a transient operation phase similar to that of a reference profile takes place so the present method can be applied.

(35) The skilled person can calculate the values of NGE and NGL from the pressure measurements, these indicators being classic.

(36) The duration of transient operating phases (in other terms the time interval during which the value of NGL or NGE is determined) is generally of the order of a few tens to a few thousands of seconds, preferably a few hundreds of seconds.

(37) In the examples of FIGS. 6b and 7a-7c, the duration of the transient operation phase is within a time interval of between 500 and 1500 seconds.

(38) In a second step (b) the data-processing unit 1 determines the value over time (over the same duration of the transient operation phase) of the Steam Range Level deviation ΔNGV, the Steam Range Level deviation corresponding to the Wide Range Level from which a component representative of a variation in free water surface in the heat exchanger has been filtered from the values of NGL and NGE.

(39) As explained, this operation is preferably the difference between the value of the NGL and the value of the NGE (ΔNGV=NGL−NGE) since the NGE is representative of the variation in free water surface.

(40) The value over time of the NGL, NGE or ΔNGV is also called respectively NGL, NGE or ΔNGV “signal”.

(41) Statistical processing will be executed conventionally and preferably by zero centering and by moving average of signals so as to standardise them (both in step (a) and in step (b)).

(42) In a step (c) the processing unit 1 compares the value of ΔNGV to a set of reference profiles ΔNGV.sub.i for said transient operation phase of the heat exchanger, each reference profile ΔNGV.sub.i being associated with a level of fouling.

(43) More precisely, there can be a database of reference profiles ΔNGV.sub.i defined for a type of transient operation phase and a level of fouling. It should be noted that there can even be different databases of reference profiles ΔNGV.sub.i associated with several types of heat exchangers. The reference profiles ΔNGV.sub.i can be stored in the memory 2.

(44) The level of fouling must be understood as a parameter representative of the extension of fouling, for example a rate between 0 and 1: a zero rate corresponds to complete absence of fouling (passage 12a, 12b completely open) and a rate of 1 corresponds to complete fouling (passage completely blocked 12a, 12b).

(45) The set of reference profiles ΔNGV.sub.i defined for a type of transient operation phase (i.e. for the whole range of fouling levels) forms a “bundle” of reference profiles such as shown in the example of FIG. 7a (for a transient operation phase of type EP RGL 4) called abacus.

(46) The lower the level of fouling, the faster the value of ΔNGV rises. In other terms the profiles at the bottom of the bundle correspond to high levels of fouling, and the profiles at the top of the abacus correspond to low levels of fouling.

(47) The reference profiles ΔNGV.sub.i can be determined empirically, in particular by digital simulation. More precisely, based on a model of the relevant heat exchanger the expected responses ΔNGV for said transient operation phase considered for each level of fouling will be precalculated. In this respect, the method advantageously comprises a previous step (a0) for generation of said set of reference profiles ΔNGV.sub.i during said transient operation phase of the heat exchanger.

(48) Developing an empirical abacus is preferably done from the knowledge of two real signals ΔNGV for a heat exchanger similar to that being considered (or even the same), for which the respective levels of fouling are known.

(49) In other terms, step (a0) advantageously comprises performing steps (a) and (b) for a reference heat exchanger similar to said heat exchanger during at least two occurrences of said transient operation phase respectively associated with a first level of known fouling and a second level of known fouling greater than the first level of fouling so as to obtain a first reference profile ΔNGV.sub.Level.sub.low during said transient operation phase of the heat exchanger for the first level of fouling and a second reference profile ΔNGV.sub.Level .sub.high during said transient operation phase of the heat exchanger for the first level of fouling, the other reference profiles ΔNGV.sub.i during said transient operation phase of the heat exchanger being calculated from the first and second profiles ΔNGV.sub.Level.sub.low and ΔNGV.sub.Level.sub.low.

(50) In practice, an empirical abacus is preferably based on a real bundle ΔNGV of a reference range, having ideally undergone cleaning, and whereof there are measurements (for example by video inspection) before and after cleaning which will respectively define the abovementioned high level and the low level, so it can be capable of covering a significant scope of levels of fouling. The extreme and intermediate levels of fouling are then extrapolated linearly from the “support” curves of the abacus.

(51) For each level of fouling considered, step (a0) comprises performing steps (a) and (b) for said reference heat exchanger similar to said heat exchanger during at least three occurrences of said transient operation phase associated with said level of fouling so as to obtain at least three real profiles ΔNGV.sub.r during said transient operation phase of the heat exchanger for said level of fouling, the obtaining of the reference profile ΔNGV.sub.i during said transient operation phase of the heat exchanger for said level of fouling comprising calculating an average of the real profiles ΔNGV.sub.r then approximation of said average by a given function.

(52) In this way, the signal ΔNGV representative of high rates of fouling (“low” abacus profile) is preferably developed by averaging the three last real signals ΔNGVr (in particular filtered and standardised) preceding cleaning, or on the date of the highest level of fouling identified over the range, then by approximating this average by a three-degree polynomial.

(53) Similarly, the signal ΔNGV representative of low levels of fouling is preferably developed by averaging the three first real signals ΔNGV following cleaning, or on the date of the lowest level of fouling identified over the range), then by approximating this average by a three-degree polynomial.

(54) These operations are illustrated in FIG. 7b which represents the polynomial approximation of the set of real signals ΔNGV.sub.r of a tranche example CA4 GV2 (only the three signals before and after fouling are in reality necessary after this). FIG. 7c represents the calculating of signals ΔNGV representative of weak and strong levels of fouling, by average respectively of the three EP after and before fouling.

(55) Once these two theoretical signals ΔNGV have been developed, two levels of fouling must be associated with them, here called Level.sub.high and Level.sub.low (Rateh.sub.high and Rate.sub.low in the preferred embodiment where the level of fouling designates a rate of fouling between 0 and 1), from the measurements available. Next, the complete theoretical abacus is developed by interpolation.

(56) Therefore, the theoretical reference signal ΔNGV.sub.i for a fouling rate Rate.sub.i can be calculated as follows:

(57) $\Delta {\mathrm{NGV}}_i=\frac{\begin{array}{c}\left({\mathrm{Rate}}_{\mathrm{high}}-{\mathrm{Rate}}_i\right).Math.\Delta {\mathrm{NGV}}_{{\mathrm{rate}}_{\mathrm{low}}}+\\ \left({\mathrm{Rate}}_i-{\mathrm{Rate}}_{\mathrm{low}}\right).Math.\Delta {\mathrm{NGV}}_{{\mathrm{Rate}}_{\mathrm{high}}}\end{array}}{\left({\mathrm{Rate}}_{\mathrm{high}}-{\mathrm{Rate}}_{\mathrm{low}}\right)}$

(58) This results in a complete abacus which can be used for estimating fouling of heat exchangers of the same type by the novel method (FIG. 7a).

(59) It will be clear however that the present method is not limited in this way to obtaining abacuses, and the skilled person can use a multitude of empirical approaches such as “machine learning” in the broad sense (deep learning, neurone networks, etc.).

(60) Returning to the method for evaluation of fouling, the result of the comparison of step (c) identifies a “target” reference profile ΔNGV.sub.opt which is that closest to the measured profile (value over time of ΔNGV obtained on completion of step (b)). The skilled person is aware of tools for identifying the most similar profile among a plurality of profiles for example by taking the difference of least squares.

(61) The level of fouling associated with the “target” reference profile ΔNGV.sub.opt constitutes a reliable estimation of the fouling of passages of the spacer plate 10 of the relevant heat exchanger, and in a step (d) this level of fouling with the identified target reference profile ΔNGV.sub.opt can be restored on an interface 3.

(62) Equipment

(63) According to a second aspect equipment is proposed such as shown in FIG. 4. It comprises a processing unit 1, a memory 2 and an interface 3. The processing unit 1 is connected to the first pressure sensor 31, the second pressure sensor 32 and the third pressure sensor 33, and is configured for executing the method according to the first aspect.

(64) A set of this equipment and of the tubular heat exchanger 11 (i.e. the steam generator), or even of the nuclear plant which comprises it, is also proposed.

(65) The invention also relates to a computer program product comprising program code instructions recorded on a carrier which can be used in a computer for performing steps for carrying out the method for evaluation of fouling, when said program is run on a computer.

(66) As a result, the pressure measurements of the sensors 31, 32, 33 are transmitted to the memory 2 to be stored there in light of its processing. This processing of measuring data to which the present invention refers is carried out by a processing unit fitted with a calculator, typically a computer provided with an interface 3, by which it acquires the measuring signal and transmits the results of executing the method for evaluation of fouling, said computer being configured to execute the method according to the invention.