Control method of the performance state of a heating system

Abstract

The object of the present invention is a method for predicting the efficiency state of a heating system comprising at least: a burner (10), at least one valve (12) for the dosing and supply of a fuel and a related combustion air supply system (13, 14), a control unit (16), at least one electrode (17) for the ignition of the air-fuel mixture and / or the detection of at least one combustion parameter, in particular an ionisation current (ION.tk). The predictive method comprises at least one measurement and / or acquisition step of at least one value of said ionisation current (ION.tk) at at least one time instant tk of an operating step of said burner (10), and a determination step of at least one "ionisation index" (ION.INDEX) and a combustion "error index" (ERROR.INDEX), the combination of said indices allowing assigning a "normal", "abnormal" or "critical" operating class to the heating system.

Description

[0001] CONTROL METHOD OF THE PERFORMANCE STATE OF A

[0002] HEATING SYSTEM

[0003] DESCRIPTION

[0004] The object of the present invention is a method for controlling the efficiency state of a heating system of the type comprising, for example, a gas boiler.

[0005] More precisely, the object of the present invention is a control method adapted to predict the ageing and / or wear and / or malfunction of one or more components of said heating system so as to promptly signal the possible need for maintenance and / or replacement interventions.

[0006] In particular, the invention falls, without any limiting intent, within the sector of gas boilers adapted to the production of hot water for room heating purposes and / or for domestic use.

[0007] Naturally, nothing prevents the method for controlling the efficiency state of one or more components of a heating system from being extended, with minimal adaptations within the reach of a person skilled in the art, to sectors similar to those of the gas boilers, for example to any other thermal energy generator obtainable from the combustion of a gas and / or liquid fuel and usable for the heating of civil, industrial and / or commercial or similar buildings and / or for the production of domestic hot water.

[0008] As an example, Fig. 1 schematically shows a typical combustion unit 1 of a heating system, in particular of a gas boiler (not explicitly shown), comprising at least:

[0009] --- one gas burner 10 housed inside a combustion chamber 11 , - one gas valve 12 for the delivery and dosing of a fuel, preferably gaseous (e.g. methane CH4, or the like), to the burner 10,

[0010] - one combustion air supply system 13 equipped with at least one fan 14, preferably with adjustable speed,

[0011] - one discharge system 15 of the gases and unburnt products generated during the combustion of the air-fuel mixture fed to the burner 10,

[0012] - one control unit 16 (or control system) capable of managing at least the usual functions of a gas boiler and components thereof.

[0013] For simplicity, the primary heat exchanger through which the heat of the flame F of the burner 10 is transmitted to a heat transfer fluid, generally water, for room heating and / or for the production of domestic hot water is not shown in said combustion unit 1.

[0014] On the other hand, 17 indicates an ionisation probe, more simply referred to as electrode 17, extended and / or in substantial proximity of the flame F of the burner 10.

[0015] Although a single electrode 17 is shown in Figure 1, it is known the possibility of providing for several electrodes, for example two or three electrodes, each assigned to one or more functions.

[0016] It is known that said one or more electrodes 17 are very important metal components in the heating systems (e.g., in the gas boilers or the like) as they are typically used to meet at least two main functions:

[0017] - the ignition, by means of a spark, of the air-gas mixture fed to the burner 10 with the consequent formation of a flame F;

[0018] - monitoring the state of said flame F both during the ignition step and during the normal operation of the burner 10, i.e. with the heating system in operating conditions.

[0019] It is in fact known the formation of ions capable of being intercepted and detected by said at least one electrode 17 of the burner 10 in the proximity of the flame F, which therefore acts as a metal collector.

[0020] Such ions, through at least one electrode 17, produce a current, referred to as “ionisation” ION, which, converted by the control unit 16 of the heating system into a digital value, hereinafter referred to as “ionisation signal”, allows controlling and monitoring the correct formation of the flame F and / or the characteristics thereof. More precisely, the detection of said ionisation current ION enables to:

[0021] - determine the formation of the flame during the igniti on step of the burner 10 and the characteristics thereof in the subsequent modulation step of the power of the heating system, and / or

[0022] - pilot and regulate, independently of one another, fuel and air supplied to the burner 10 so as to ensure, at least in the boilers having electronic combustion control systems (CCS), an optimal combustion and compensate for possible variations in the quality of the gas fed and / or the drift of some components of the heating system.

[0023] The detection and analysis of the ionisation current further allows monitoring the performance and reliability of the heating system and the main functional components thereof.

[0024] Fig. 2 shows a typical time (or profile) trend of the ionisation current ION detected by an electrode.

[0025] The value of the ionisation current ION is shown on the ordinate axis, while the abscissa axis represents the time.

[0026] During the ignition step, in particular during the very first instants, the striking of the spark of the electrode 17 and the triggering of the combustion of the air-fuel mixture supplied to the burner 10 causes a sort of “small explosion” in the combustion chamber 11 and the ionisation current ION reaches a first peak, referred to as lON.tpeak.

[0027] Starting from said peak, the ionisation current ION starts to progressively decrease until reaching a minimum point, referred to as lON.tvalley.

[0028] Such effect, known in jargon as the “valley effect”, is generally due to a lower capacity of the electrode to detect the ionisation current ION (increase in the equivalent resistance of the electrode), e.g. depending: - on the oxide state formed thereon (such aspect shall be extensively referred to during the present description) and the cold conductivity thereof, and / or

[0029] - on specific functional / operating conditions of the burner 10. Subsequently, once the minimum has been reached, due to the increase in the temperature of the electrode, the conductivity of the oxide layer will tend to increase and the ionisation current ION will again begin to increase with a slower rate relative to the very first instants of the ignition step, up to stabilising around an ignition-end value. The term t.safety indicates a safety time of the ignition step at which it is evaluated whether the flame was formed correctly or not.

[0030] In the absence of a flame, or if a rather weak and / or unstable flame F is detected, the boiler may generate one or more warning notifications and proceed with further ignition attempts, without prejudice to the possibility of providing a safety lockout if said failed and / or anomalous ignitions should repeat for an excessive number of consecutive times.

[0031] Conversely, if at instant t.safety a strong and stable flame F is detected, the ignition step may be considered as completed and the heating system may switch to a subsequent operating mode, referred to as “modulation” or “calibration”, during which the power of the burner 10 is suitably regulated to ensure the user the right heating comfort.

[0032] More precisely, the modulation or calibration step, however available only for the heating systems (e.g. boilers) operating and implementing an electronic combustion control system (CCS), guarantees the correct value of λ (ratio between the air and fuel with respect to the stoichiometric ratio of the fuel used), i.e. the right and balanced supply of fuel and air to the burner 10.

[0033] It is also known that any defect and / or deterioration of one or more of the components of the combustion unit 1 of a heating system may lead to errors in the detection of the flame F and, as already anticipated, to consequent possible safety lockouts. Such situation is a cause of disservice for the user who must independently provide for clear the error, e.g. via the HMI interface of the heating system, an operation that is not always simple and intuitive for an inexperienced user, and / or contact the technical assistance for a maintenance intervention of the same boiler.

[0034] In particular, it has been observed that during the operational life of a boiler, it is important that the performance of an electrode is affected as litle as possible by the natural and inevitable ageing and / or that the distance thereof from the burner remains substantially the same, in order to have good and repeatable ignitions, and therefore reliable detections and monitoring of the flame F.

[0035] However, this is not always possible.

[0036] For example, the thermal stresses to which said at least one electrode is subject during the life of the boiler may lead to significant deformations and flexions thereof with a consequent possible increase in the distance thereof from the burner 10; this may lead to a worsening of the interaction with the flame F and a worsening of the detection of the ionisation current ION.

[0037] It is clear that an excessively deformed electrode must necessarily be replaced with a new one, so as to ensure operation continuity of the heating system.

[0038] It is also known that, with the succession of the combustions, the aforementioned oxide layer is formed on the external surface of the electrode which may compromise the detection of the ionisation current and the subsequent conversion thereof into an ionisation signal which may be used to predict and signal any anomalies or heating system fault conditions.

[0039] The formation of said oxide layer mainly concentrates in the first hours (usually within the first 500 hours) of operation of the burner 10, to then progressively slow down and stabilise.

[0040] Consequently, the efficiency of the detection of the ionisation current ION detected is not proportional to the “operation” time (or, equivalently, “ageing time”) of the heating system which is, for example, quantifiable in weeks, months or years. As schematically shown in the graph of Fig. 3, which provides for the operating time of the heating system on the abscissa and the absolute values of the ionisation current detected from time to time during the ignition of the burner 10 on the ordinate, the following are identified: - a first part “A” in which the electrode is new and the absolute values detectable of the ionisation current ION are high, although subject to an initial rapid decrease due to the aforementioned “substantial” oxidation phenomena (see, for example, the negative trend with high slope represented by the line A), - a second part “B” in which an increase in the oxide layer in said electrode corresponds to a further decrease in the ionisation current ION detected, although with a slower speed and trend than in the first part A (see, in this regard, the “less sloping” trend of the strai ght line B),

[0041] - a third part “C” in which the ageing and oxidation effect of the electrode stabilises reaching a condition that may be defined as essentially stationary

[0042] (see the straight line C with a very low slope). In this case, i.e., due to the effect of the oxidation, the ionisation current detectable by an electrode will settle on very low absolute values, which shall remain substantially constant or slightly and slowly decreasing, as long as the heating system continues to operate.

[0043] When the values of the ionisation current ION, detected during the ignition step of a burner, reach and / or fall below a safety threshold, shown in Fig. 3 with the horizontal line “LS”, there is a high risk of a failed or unstable ignition of the flame F, with the consequent signalling of at least one warning or notification of such adverse event and / or lockout of the heating system.

[0044] Various interventions are possible in order to make up for the formation of the oxide layer on at least one electrode of a heating system, and avoid or limit the relative inconveniences and malfunctions; for example, in addition to a radical replacement thereof, it is possible to provide for periodic maintenance operations for the removal, by brushing, of the oxide layer. Alternatively, especially when the ageing and oxidation degree of the electrode has not yet reached undesirable levels, it is possible to adjust the lambda λ of the air-fuel mixture fed to the burner 10, for example by enriching it, so as to compensate for the low values of the ionisation current ION that said electrode would be able to detect, favouring the acquisition thereof.

[0045] However, it is not always simple and intuitive to adopt one of the above intervention and / or maintenance measures correctly and at the right time.

[0046] For example, it has already been said that in stationary conditions (segment C of the graph in Fig. 3), the values of the ionisation current ION and the related ionisation signals deriving therefrom have a substantially stable trend, slowly reducing over the ageing and / or functioning time of the heating system.

[0047] This may lead to the risk of providing and supposing critical operating conditions of the heating system too early or with an unacceptable delay; in the first case there would be a negative impact on maintenance costs running the risk of substituting and replacing a component that is still sufficiently operational, while in the second case malfunctions of the heating system or sudden safety- lockouts thereof would not be foreseen in time, with consequent inconveniences for the user.

[0048] Furthermore, the ignition of a burner is a transient and critical step during which the measurement of the ionisation current and / or the detection of the relative ionisation signals is complex and not perfectly repeatable.

[0049] The ageing and oxidation of an electrode are not the only factors that may affect the operation and performance of a heating system upon ignition, it is known, for example, that for a correct and stable formation of the flame of at the burner, it is also necessary that specific arrangements of design and / or installation and / or functional nature are substantially met. By way of an example, it is necessary that the most correct geometric configuration and length of the flue of the heating system for the discharge of unburnt products and combustion fumes is chosen and / or to carefully consider the presence of recirculation of said fumes and / or the supply pressures of fuel gas at the burner. In the light of all of the above, it therefore appears obvious that the mere detection of an ionisation current, as provided for in various documents of the state of the art (see, by way of an example, the prior patents DE 100 03 819 C1 and / or DE 100 57 233 A1), may not be sufficient to predict, sufficiently in advance, problems and malfunctions of the heating system or, least of all, provide information on the probable cause thereof.

[0050] The methods for monitoring and determining the ageing of one or more components of a heating system described in the aforementioned state-of-the-art documents also have further problems and limitations.

[0051] For example, in the document DE 100 03 819 C1, for the determination of the ionisation current, the burner is deliberately operated at different power ranges, which results in undesirable energy consumption and an increase in pollutants.

[0052] In this document, it is also provided that the ionisation current is determined in at least two different instants for each of the powers at which the burner is operated (for example, said burner may be operated at rated load and at partial load). This also leads to an undesirable extension of the monitoring times.

[0053] The method described in the prior document DE 100 57 233 A1 instead, while monitoring the ionisation current only during the ignition step of a burner, and therefore for a more limited period of time, however requires the acquisition and exact knowledge of the entire time profil e thereof. This results in a complication of the same monitoring method, and of the relative electronics adapted to implement it.

[0054] Finally, other known methods provide that the monitoring of said ionisation current is carried out during specific trial (or test) steps of the burner, during which the normal operation and use thereof is therefore temporarily suspended. The object of the present invention is to obviate such kind of drawbacks by providing an innovative method for predicting the performance state of a heating system based on the monitoring of a combustion parameter, for example of at least one ionisation current.

[0055] More precisely, the object of the present invention, at least for some executive embodiments thereof, is to provide for a method for predicting the performance state of a heating system able to notify the need for an intervention and / or maintenance at the right moment, for example in advance relative to the risk that said heating system reaches a lockout condition.

[0056] A further object of the present invention, at least for some executive embodiments thereof, is to provide for a method for predicting the performance state of a heating system able to provide information on the causes of possible malfunctions and / or wear of said one or more components.

[0057] A further purpose of the present invention is to provide for a method for predicting the performance state of a heating system which does not require the knowledge and acquisition of the entire time profile of a combustion parameter, for example of an ionisation current, during at least the ignition step of the same heating system.

[0058] A further object of the present invention is to provide for a method for predicting the performance state of a heating system which may be implemented without the need to resort to specific test steps which provide for the temporary interruption of the normal operation of the same heating system and / or excessive and additional energy consumptions and polluting emissions.

[0059] A further object of the present invention is to provide for a method for predicting the performance state of a heating system that is simple to implement in the electronics already provided in known heating systems and which is quick to be implemented and executed.

[0060] These and other objects, which shall become clear below, are achieved with a method for predicting the performance state of one or more components of a heating system in accordance with the provisions of the independent claims.

[0061] Other objects may also be achieved by means of the additional features of the dependent claims.

[0062] Further features of the present invention shall be better highlighted by the following description of a preferred embodiment, in accordance with the patent claims and illustrated, purely by way of a non-limiting example, in the annexed drawing tables, wherein:

[0063] - Fig. 1 schematically shows a heating system according to the state of the art in which the method for predicting the efficiency state of one or more of the components thereof according to the invention may be implemented,

[0064] - Fig. 2 shows a graph representative of the time profile of a combustion parameter, for exampl e of an ionisati on current, during an ignition step of a heating system;

[0065] - Fig. 3 shows a graph representative of a change over time in a combustion parameter, for example of an ionisation current, at a particular instant of an ignition step of a heating system;

[0066] - Fig. 4 shows a graph representative of at least one step of the method for predicting the performance state of a heating system according to the invention

[0067] - Fig. 5a shows a graph for identifying an operating class of said heating system;

[0068] - Figs. 5b-5c respectively show two classification examples of the operation of said heating system;

[0069] - Figs. 6a-6b show further graphs representative of some steps of the method for predicting the performance state of a heating system, according to the invention.

[0070] The features of a preferred variant of the method for predicting the efficiency state of a heating system according to the invention are now described using the references contained in the figures. Hereinafter, for summary and clarity of description, said method for predicting the efficiency state of a heating system will be simply referred to as “predictive method”.

[0071] Said predictive method is suitable for being applied to any known heating system, in particular to any generator of thermal energy obtainable from the combustion of a gaseous and / or liquid fuel, such as, for example, gas boilers for heating civil, industrial and / or commercial buildings, or the like and / or for the production of domestic hot water. For the purposes of the present description, reference shall be made, without any limiting intent, to a gas boiler comprising a combustion unit 1 of the type schematically illustrated in Fig. 1 and already widely described.

[0072] As anticipated, only the main components necessary for the implementation and execution of the predictive method of the invention are shown of said combustion unit 1, such as at least the burner 10, the one or more electrodes 17, etc., which shall be referred to if necessary, from time to time, during the present description.

[0073] For the purposes of the invention, as already partly mentioned, the control unit 16 of the heating system is able to manage the operation of said burner 10 and / or other components cooperating therewith (gas valves 12, fan 14, electrode 17, etc.), with which it may “communicate” bidirectionally according to the most appropriate methods, for example, via wired and / or wireless connections (radio type, such as WLAN, ZigBee, Bluetooth protocols or the like).

[0074] Naturally, nothing prevents one or more of the functions managed by said control unit 16 from being partially or entirely carried out on devices external to the heating system, such as, by way of a non-limiting example, smartphones, tablets, PDAs, remote servers (e.g. in Cloud), as long as they are suitably equipped with software and / or applications suitable and compatible for the purpose.

[0075] Obviously, said control unit 16 and / or the possible external devices may be further conceived and arranged to implement and manage one or more steps of the method for predicting the performance state of a heating system according to the invention.

[0076] In other words, said control unit 16, which, as mentioned, may be the same control system that subtends the “general” operation of the heating system, an ad hoc external control system integrated / integrable with it (e.g. a retrofit kit), or a remote processing logic, e.g., implemented in external mobile and / or Cloud devices, is, without any limiting intent advantageously provided and / or cooperating with: - input data acquisition and / or receiving means for the predictive method of the invention, said input data being able to comprise, as it shall be seen, at least one combustion parameter,

[0077] --- means for calculating and processing said input data adapted to provide output data and / or information aimed at determining the efficiency state and condition of the heating system,

[0078] - memory means for the at least temporary storage of said input and / or output data,

[0079] - means for transmitting said output data and / or information towards a display and notification interface and / or for managing one or more steps of said predictive method of the invention, said interface being able to consist of a typical HMI interface integrated in the heating system and / or the same external device mentioned above.

[0080] Said predictive method, like various known solutions of the state of the art, is essentially based on the monitoring of a combustion parameter, preferably of at least the ionisation current which, as known:

[0081] --- is generated when the ions produced by the combustion at the burner 10 of a liquid or gaseous fuel hit the relative at least one electrode 17,

[0082] --- may be converted into a digital value, i.e. into an “ionisation signal”, via, for example, the control unit 16 mentioned above.

[0083] Hereinafter, for clarity of description, the term “ionisation current” shall mean the same current and / or, indifferently, the ionisation signal derived therefrom.

[0084] More precisely, such predictive method aims to measure and control the ionisation current during an operating step of the burner 10.

[0085] Unlike the solutions of the state of the art which provide for the knowledge of the entire time profile of the ionisation current, the predictive method of the invention provides for the measurement and acquisition of at least one value of the ionisation current lON.tk, preferably only one value assumed in a single time instant tkduring an operating step of the burner 10.

[0086] According to the invention, said measurement and acquisition of a value of the ionisation current lON.tkis then carried out and / or repeated, according to same methods, one or more times within an appropriate “observation” time window T.analysis of the operating life of the burner 10.

[0087] Preferably said observation window T.analysis, may be of a predetermined duration, for exampie equal to one day, one week, one month or one year, as well as multiples or fractions thereof.

[0088] Preferably (but not necessarily, as shall be seen), said predictive method aims at measuring said ionisation current lON.tkat an instant tkduring the initial ignition step of the burner 10.

[0089] Preferably, the measurement and acquisition of said specific value of the ionisation current lON.tkmay be repeated for each of the i-th ignitions of the burner 10 found in said observation time window T.analysis .

[0090] According to an executive embodiment of the invention, which is among the preferred ones, said specific value of the ionisation current lON.tk, acquired and stored at the instant tkfor each of said ignitions “i” of the burner 10 within an observation time window T.analysis, may be measured at the “safety time” t.safety (tk= t.safety), i.e. substantially at the end of said first ignition step; in fact, it shall already been seen that t.safety represents a default safety time (e.g., chosen during the design and / or installation-commissioning step of the heating system) at which a flame should generally have already been formed and stabilised.

[0091] Said value of the ionisation current at the safety time t.safety is also referred to as “flame value lON.tflame'’ (lON.tk= lON.tflame).

[0092] Naturally, as an alternative, nothing prevents the possibility of taking into account, with minimal adaptations within the reach of a person skilled in the art, other significant values of the ionisation current lON.tkat a specific instant tkduring the ignition of the burner 10 and prior to the safety time t.safety, for example the maximum flame value ION.tpeak, the minimum value lON.tvall ey(valley effect), the ignition time value lON.tignition(with t.ignition < t.safety), or combinations thereof. Preferably, said predictive method is designed to consider and select among the measured and / or acquired ionisation current values lON.tk, those representative of “cold” ignitions of the heating system, i.e. occurred after a “cooling time” t. cooling of the burner 10, during which it remained inactive and / or the related at least one electrode 17 not exposed to any flame F.

[0093] In fact, it is known that during a cold ignition of a burner 10, the increase in the resistivity of the at least one electrode 17 is, for example, more sensitive and evident, due to the oxide layer formed on the surface thereof, that making more significant the corresponding values actually detected of the ionisation current lON.tk.

[0094] For example, in such conditions, the oxide layer possibly present may amplify the "‘valley effecf with the value of the ionisation current lON.tvallevwhich may reach very low values and with the consequent risk of failed or incomplete ignition of the flame F at the time t.safety.

[0095] Therefore, considering the values of the lON.tkas representative of cold ignitions may facilitate and improve the determination of the efficiency / performance state and “health” of the electrode 17 and / or of the combustion.

[0096] The predictive method of the invention also comprises a routine (or logic) capable of applying different operations and / or mathematical processing to the set of values of the ionisation current lON.tkmeasured and / or acquired for each observation time window T. analysis and of identifying and considering any combustion or ignition errors so as to recognise, and possibly signal, a specific operating condition of the heating system.

[0097] More specifically, according to a possible executive embodiment of the invention, which is among the preferred ones, said routine is aimed at extrapolating from said values of the ionisation current lON.tkat least a pair of “indices”, whose values, when suitably combined to one another, (as shall be seen), enable to assign the monitored heating system an operating class: “normal”, “abnormal” or “critical”, each representative of a specific efficiency / pertormance state of the same heating system and / or the presence of possible problems or disservices thereof.

[0098] However, according to further variants of the invention, nothing prevents the determination of at least one of said indices from depending and / or deriving from characteristic parameters of a flame at the burner 10 alternative to the ionisation current lON.tk. Such aspect shall be referred to extensively hereafter during the present description.

[0099] It should be noted that, while “normal” operation does not produce any concern or signalling, an “anomalous” operation indicates an operating situation of the heating system still substantially acceptable but which, in the short / medium term, may degenerate, while a “critical” operation corresponds to a situation that is already substantially compromised.

[0100] An “abnormal” or “critical” operation found and detected for several consecutive observation time windows T.analysis is interpreted by the predictive method of the invention as representative, respectively, of a potential and / or already existing problem of the heating system to be notified to the user (and / or to the assistance sendee).

[0101] More precisely, the predictive method of the invention may produce a notification of:

[0102] - “combustion check”, if an “abnormal” operation is detected for N consecutive observation time windows T.analysis,' and / or

[0103] - “electrode check”, if a “critical” operation is detected for M^ consecutive observation time windows T.analysis, with N, M previously chosen equal to 1 , 2 ... n.

[0104] Preferably, N may be chosen greater than M (N>M).

[0105] Such signalling or notification may be of a visual and / or acoustic type, for example shown in the HMI interface of the heating system and / or of any external devices connected thereto and / or by written communications via email, SMS, or the like, to the same external devices or remote servers.

[0106] Without any limiting intent, a “combustion check” notification may concern a state of potential and / or imminent malfunction of the heating system attributable to combustion defects, caused, for example, by inaccurate calibrations and / or adjustments of the combustion parameters, and / or by installation and commissioning problems of the same heating system.

[0107] In this case, the predictive method of the invention may also notify and suggest the possible factors to be controlled and arranged, for example the configuration / geometry and length of the flue for the discharge of unbumt products and combustion fumes, their recirculation, the combustion ratio λ, etc. in order to help the installer and / or the technical assistance and speed up the operating recovery intervention (through which the heating system may again switch from said “abnormal” to the “normal” operating class).

[0108] On the other hand, the “electrode check” notification preferably concerns a malfunction condition of the monitored heating system attributable to the combustion unit I and linked, in general, to an excessive oxidation of the external surface and / or to significant deformations and / or to a definitive breakage of the electrode 17 and / or to a breakage of the burner 10.

[0109] Similarly, even in case of “electrode check”, the predictive method of the invention may possibly associate, with the notification provided, useful information for solving the problem, by signalling, for example, a radical replacement of the faulty electrode 17 and / or the need for a maintenance intervention for the removal, by brushing, of the oxide layer.

[0110] The “indices” adapted to classify and associate the heating system to one of said operating classes shall be extensively referred to during the present description.

[0111] However, it is already necessary' to specify that, according to a first possible executive embodiment of the invention, and as partially anticipated, from the set of values of the ionisation current lON.tk, measured and / or acquired in an observation time window T.analysis, only those most significant values are selected and extrapolated to assign the most appropriate operating class to the monitored heating system; for example, the representative values of:

[0112] - ignitions, preferably cold, of the burner 10, adapted to define an index ION.INDEX referred to as “ionisation index”, and

[0113] --- combustion errors of the burner 10, adapted to define an index ERROR.INDEX referred to as “combustion error index” (or, more simply, abbreviating, “error index”). Preferably, the determination of said indices ION.INDEX and ERROR.INDEX may be carried out with reference to the same observation time window T. analysis, and repeated substantially identical for each of the observation time windows T.analysis provided during the entire monitoring period of the heating system. According to one embodiment of the invention, which is among the preferred ones, said ionisation index ION.INDEX is a determined value, initially:

[0114] - by filtering from said acquired values of the ionisation current lON.tkonly those corresponding to:

[0115] - ignitions of said burner 10 taken place after a suitable cooling time t.cooling which, as anticipated, represents a suitable time interval, between two consecutive ignitions, during which the burner 10 remained substantially inactive, for example, t.cooling may be preferably equal to 30 minutes; and / or,

[0116] - at internal temperatures of the heating system lower than a predetermined threshold (e.g., ≤ 35°C) and detected, e.g., in the proximity of or at the same burner 10, provided for this purpose with at least one suitable temperature sensor (not shown) or the like, and then, subsequently

[0117] - by applying further mathematical processing on said filtered and residual values, of the ionisation current lON.tkcomprising at least one operation which may consist of:

[0118] - calculating a specific absolute value thereof, for example, the maximum, the minimum and / or an average (arithmetic or weighted), and / or - calculating the trend or slope of a curve interpolating said residual values, and / or

[0119] --- carrying out statistical or similar processing or the like, for example, by calculating the 90th percentile or the like (which, as known, represents a value below which a given percentage of the other residual values “falls within”).

[0120] As an example, on the graph of Fig. 4, which predicts the operating time t of a burner 10 on the abscissa within an observation window T.analysis and the ionisation indices ION.INDEX on the ordinate, the values of the ionisation current lON.tkare shown (in the specific case the flame value lON.tflameat the safety time t.safety), wherefrom, an ionisation index ION.INDEX equal to 82 points is graphically determined according to one of the processing listed above or combinations thereof; in such regard, see the dashed “trend or tendency line” L-L.

[0121] With reference, on the other hand, to the determination of the aforementioned error index ERROR.INDEX, the combustion errors of the burner 10 may be preferably taken into account from the predictive method of the invention, able to cause at least temporary lockouts of the heating system and dependent, for example, on malfunctions of specific components and / or combustion defects thereof. Such data is in fact easy to find and handle, because the current and known control units 16 of a heating system (for example of a gas boiler) are already normally arranged to signal such type of errors.

[0122] I.e. the predictive method of the invention may consider any failed or insufficient activation of the flame F as “combustion errors” at the burner 10 (i.e. a substantial flame absence) at the end of each ignition attempt “i” occurred within an observation time window T.analysis.

[0123] In fact, it is known that the flame absence F, for example at the safety time t.safety, is easily identifiable since no ionisation current lON.tkis detected by the at least one electrode 17 of the burner 10. Similarly, a weak and / or unstable flame F, although generated, may be coded as a “combustion error” if the values of the ionisation current lON.tkdetected are lower than the default safety thresholds, possibly pre-stored in said control unit 16.

[0124] As anticipated, in alternative variants of the predictive method of the invention, said combustion errors may be determined independently of the values of the ionisation current lON.tk, for example by arranging the heating system 1 with special sensors capable of detecting and / or directly evaluating characteristic parameters of the presence or absence of a flame at the burner 10, to be compared with the reference threshold values.

[0125] Without any limiting intent, said characteristic alternative parameters may comprise, for example: the intensity of the visible, infrared and / or ultraviolet light radiation of a flame, the temperatures reached in the proximity or at the same burner 10, the noise associated with the combustion, or similar parameters. As a result, said sensors, also known as “flame sensors”, may comprise known optical, temperature, ultrasound sensors, or similar / equivalent devices able to detect whether or not a flame has formed correctly at the burner 10.

[0126] The predictive method of the invention is therefore able to extrapolate, by means of specific and / or known software or algorithms, said error index ERROR.INDEX from all the combustion errors identified and stored by the control unit 16 within an observation time window T.anafysis (which, as mentioned, may be the same as the determination of the ionisation index ION.INDEX, or possibly of longer duration), regardless of the method adopted for their detection.

[0127] For simplicity of description, in accordance with the examples described up to now, said “combustion errors” may also be considered and indicated as “ignition errors”.

[0128] Without any limiting intent, said error index ERROR.INDEX may be representative of the flame absence frequency (unstable and / or incorrectly formed flame).

[0129] This error index ERROR.INDEX may therefore take an absolute value comprised between 0 and 1, where:

[0130] - “0” (zero) indicates that no “ignition” errors have occurred within the observation time window T.analysis,'

[0131] - “1” indicates a high frequency of ignition errors distributed in the same time window T.analysis.

[0132] It is useful to further specify that, in the event of failed or incomplete ignition of the flame F, the control unit 16 generally attempts several times, preferably at least twice, to activate the burner 10 before the heating system is switched to a security lockout state. This means that said control unit 16 may produce at least two warnings before notifying an “ignition error” and stopping the heating system; typically, a typical error sequence may provide for: a “1st warning → 2nd warning error notification → possible system lockout” .

[0133] Therefore, when calculating the error index ERROR.INDEX the routine or algorithm implemented by the method of the invention may be programmed for:

[0134] - knowing and considering the expected error notification sequence for the specific monitored heating system, and

[0135] - assigning a different importance to each of the warnings provided for before the lockout, by assigning, for example, a greater importance to the second warnings, following which, as a rule, the error notification occurs since the risk of heating system lockout is more imminent and likely to occur.

[0136] As already said, the combination of the values of the ionisation index ION.INDEX and of the error index ERROR.INDEX allow the operation of the monitored heating system to be associated with one of the “normal”, “abnormal” or “critical”, operating classes listed and described above.

[0137] This is graphically shown in the graph of Fig. 5a which shows the values of the ionisation indices ION.INDEX on the abscissa and the error indices ERROR.INDEX on the ordinate, and / or, more precisely, in the graphs of Figures 5b and 5c referring, for example and without loss of generality, to the monitoring of a heating system in two different periods of the operating life thereof and wherein each point defines the combination of said two indices calculated and defined in an observation time window T.analysis.

[0138] Preferably:

[0139] - the “normal” operating class CL.N may be substantially characterised by at least:

[0140] - a substantial absence of ignition errors (ERROR.INDEX approximately or equal to 0), regardless of the value of the ionisation index ION.INDEX, and / or

[0141] - low or very low values of the error index ERROR.INDEX and predominantly medium or high values of the ionisation index ION.INDEX;

[0142] - the “abnormal” operating class CL. A is substantially characterised by medium or high values for both the error index ERROR.INDEX and the ionisation index ION.INDEX,

[0143] - said “critical” operating class CL.C is substantially characterised by at least medium or high values of the error index ERROR.INDEX and by low values of the ionisation index ION.INDEX.

[0144] It should also be noted that low or very low values of both the error index ERROR.INDEX and the ionisation index ION.INDEX may sometimes represent both a “normal” and “critical” operation; in such regard, see, for example, the dashed area AT in the proximity of the intersection of the axes of the ordinates and abscissas of the graph in Fig. 5a -5c.

[0145] In fact, very low values of the ionisation index ION.INDEX may sometimes denote a particularly unstable and uncertain operation of the heating system. In such case, even small variations and reductions in the value of the ionisation index ION.INDEX, the error index ERROR.INDEX being equal (or vice versa), may suddenly lead to classify the heating system from a “normal” operation to an already substantially “critical” operation, or vice versa.

[0146] This may make more complicated to assign the heating system the most correct operating class between “normal” or “critical” one, with the risk of underestimating a possible problem or, on the contrary, becoming “overalarmed”.

[0147] To overcome such problem, as shown in Fig. 5a-5c, it was preferred to introduce safety thresholds for said ionisation ION.INDEX and error indices ERROR.INDEX capable of identifying the areas capable of defining with greater precision the aforementioned operating classes CL.N, CL. A, CL.C of the monitored heating system.

[0148] In general, the choice of said thresholds may depend and vary / according to the technical and functional features of the monitored heating system.

[0149] For example, from laboratory / tests, for the ionisation index ION.INDEX it was preferred to consider a first limit threshold LIM.1, referred to as “flame detection”, and a second upper threshold LIM.2 (with LIM.2 > LIM.1), both represented by the two vertical dashed lines, while for the error index ERROR.INDEX the threshold LIM.3 was considered.

[0150] Preferably, as shown in Fig. 5a, said threshold LIM.3 of the error index ERROR.INDEX may consist of a first segment LIM.3a, substantially horizontal, and of a second pending segment LIM.3b which bends downwards (i.e. towards the abscissas), e.g. with a linear (as shown) or parabolic pattern.

[0151] More precisely, said threshold LIM.3 remains substantially constant (segment LIM.3a) at the medium or high values of the ionisation index ION.INDEX while it bends downwards (segment LIM.3b) for low, or very low values, of the same index ION.INDEX (for example, for values comprised between the thresholds LIM.1 and LIM.2).

[0152] In such way, as already mentioned above, the most correct operating class between the “critical” or “normal” one may be assigned to the heating system even in presence of low or very low values of the ionisation index ION.INDEX and of the error index ERROR.INDEX, making the predictive method of the invention more efficient and accurate.

[0153] In fact, in the light of what illustrated just above: - the “normal” operating class CL.N may be defined between said threshold line LIM.3 of the error index ERROR.INDEX and the abscissas axis,

[0154] - the “abnormal” operating class CL. A is identified between said second threshold LIM.2 of the ionisation index ION.INDEX and the horizontal segment LIM.3a of said threshold LIM.3 of the error index ERROR. INDEX,

[0155] - the “critical” operating class CL.C is delimited by said first and second threshold LIM.1 and LIM.2 of the ionisation index ION. INDEX and by the sloping segment LIM.3b of said threshold LIM.3 of the error index ERROR.INDEX.

[0156] Even more precisely, by way of a non-limiting example, said:

[0157] - “normal” operating class CL.N may be defined by values of ION.INDEX

[0158] ≥ LIM.1 and of ERROR.INDEX ≤ LIM.3,

[0159] - “abnormal” operating class CL. A may be defined by values of ION.INDEX ≥ LIM.2 and ERROR.INDEX ≤ LIM.3a.

[0160] - “critical” operating class CL.C may be defined by values of LIM.1 ≤ ION.INDEX ≤ LIM.2 and ERROR.INDEX ≥ LIM.3b.

[0161] As already partially anticipated, if an “abnormal” operation is detected for N consecutive observation time windows T.analysis, with N for example equal to 3 (N=3), the predictive method of the invention notifies, in the display and / or communication modes mentioned above, the “combustion check”.

[0162] Similarly, if a “critical” operation is detected for M consecutive observation time windows T.analysis, for example with M equal to 2 (M==2), the “electrode check” notification is generated.

[0163] In general terms, it is advisable that, for an operating state already considered as “critical”, the relative “electrode check” notification is transmitted and / or displayed quicker than the “combustion check” notification times (as mentioned, preferably after two consecutive T.analysis, instead of the three generally- provided in case of “anomalous” operation classification).

[0164] This is because in “critical” conditions the operation of the heating system is already substantially compromised (the system may already be locked out) and therefore it is advisable to plan a consequent maintenance or restoration intervention with the maximum urgency.

[0165] It therefore appears clear that, depending on the notified “check”, the predictive method of the invention enables to take into account and report not only possible malfunctions of the heating system deriving from the malfunction, wear and / or breakage of one or more of the components thereof, for example of the electrode 17, but also from other different factors and phenomena, such as problems and defects relating, for example, to the combustion and / or to the design / installation of the same heating system and / or to the operating and / or environmental conditions in which it operates (for example, cold ignitions, incorrect calibrations of the combustion, unsuitable technical and geometric features of the flue for discharging the combustion fumes, particular atmospheric conditions, e.g., very windy, etc.).

[0166] This results in into a more effective evaluation of the efficiency condition or excessive ageing reached by a heating system, so as to plan, in the right times, maintenance and / or restoration interventions on one or more of the functional components thereof.

[0167] By way of an example, and without any limiting intent, figures 6a and 6b show a possible monitoring of the performance state over time of a laboratory-tested heating system and implemented by the predictive method of the invention.

[0168] In both graphs of Figures 6a and 6b, which show the operating time of a heating system on the abscissa axis and the values of the ionisation current on the ordinate axis:

[0169] - the points Pi represent the values of the ionisation current ION.tkmeasured and / or acquired, for example, according to a possible embodiment of the invention, at the safety time t.safety (lON.tk= lON.tflame) and for each of the i-th ignitions, preferably “cold”, of the burner 10 found within a generic monitoring period of the heating system, comprising at least one observation time window T.analysis,' - the continuous vertical lines indicate the “combustion / ignition errors” reported by the control unit 16 of the monitored heating system within an observation time window T.analysis,'

[0170] - the values “c” and “e” represent, respectively, the ionisation index ION.INDEX and the error index ERROR.INDEX which may be derived, in the ways illustrated above, starting from the aforementioned values Pi of the ionisation current and from the combustion / ignition errors.

[0171] Based on a combination of such indices, the heating system is assigned a specific “normal” CL.N, “abnormal” CL. A or “critical” CL.C operating class which may also be visually represented by assigning a different background colour with more portions of the plane of Figs. 6a and / or 6b (e.g., a green background for a “normal” operation, a yellow / orange background for an “abnormal” operation, a red background for a “critical” operation); - finally, the dashed vertical lines represent the notification of “combustion check” or “electrode check” generated by the predictive method of the invention following the permanence of the heating system in one of the “abnormal” or “critical” operating classes for several consecutive T.analysis (N, M consecutive permanencies, respectively). For example, without any limiting intent, the observation time windows T.analysis in the graphs of Fig. 6a and 6b have been chosen for a duration substantially equal to one week each.

[0172] Fig. 6a shows the case of a “combustion check” notification and by analysing it in more detail it may be seen that: - in WEEK_ 1 and WEEK_ 2, the ionisation index ION.INDEX is medium- high (“c” equal to about 66-67 points) while the error index ERROR.INDEX is low (“e” equal to about 0.15 i.e. low frequency of failed ignitions at t.safety); consequently, in this monitoring time interval, comprising one or more observation windows T.analysis, the predictive method assigns a “normal” operating class to the heating system, as also shown by the relative points W1, W2 in the graph of Fig. 5b;

[0173] - starting from WEEK_ 3 onwards, instead, even if the ionisation index ION.INDEX remains substantially constant on medium-high values (about 65-67 points), there is a substantial increase in the error index ERROR.INDEX towards average values (i.e. on the order of 0.30-0.40).

[0174] Consequently, in this monitoring time interval comprising one or more observation time windows T.analysis, the heating system switches from the “normal” to the “abnormal” operating class as defined by the points W3 ... Wm of the graph of the same Fig. 5b; - starting from WEEK_ 6, the predictive method of the invention generates a

[0175] “combustion check” notification having been ascertained that the monitored heating system remained within the “abnormal” operating class for N=3 consecutive observation time windows T.analysis (WEEK_3 WEEK 5). Similar reasoning may also be extended to the case of an “electrode check” notification, shown in Fig. 6b.

[0176] In this case, by way of an example and without any limiting intent:

[0177] - for the first four monitoring weeks (from WEEK _1 to WEEK_ _4), low values of the ION.INDEX ionisation indices are observed (on the order of 20-25 points) but, at the same time, low values of the error index

[0178] ERROR.INDEX (even equal to 0 in the example; no continuous vertical line) and therefore a substantially “normal” operation of the heating system; see, in this regard, points W1, W2, W3, W4 in the graph of Fig. 5c and what described with reference to the particular “descending” profile of at least the segment LIM.3b of the relative threshold line LIM.3,

[0179] - from WEEK_ 5 onwards, substantially constant and low ionisation index values ION.INDEX are observed (the ionisation current signal flattens out) but medium-high error index values ERROR.INDEX (in some cases even higher than 0.5); this corresponds to the passage of the monitored heating system from a “normal” operating class directly to a “critical” operating class, graphically displayed in Fig. 5c from points W5 . . . Wm,

[0180] - starting from WEEK_ 6, the predictive method of the invention generates an “electrode check” notification having been ascertained that the monitored heating system remained within the “critical” operating class, as already said by way an example, for M=2 consecutive observation time windows T. analysis (WEEK_5 and WEEK_6).

[0181] As anticipated, for the sake of clarity, it should again be noted that in both examples of Fig. 6a and 6b, also for simplification reasons, the observation windows T.analysis have been considered preferably of substantially weekly duration so as to coincide with the aforementioned weeks WEEK_m for monitoring the operation of the burner 10.

[0182] Naturally nothing prevents, as seen, these observation windows T.analysis from taking different durations from the weekly one, e.g. daily or monthly or yearly (or fractions or multiples thereof) within the operating life of the burner 10, this influencing the number of “combustion check” or “electrode check” notifi cations generated when the aforementioned conditions occur.

[0183] In general, for a more efficient use of the predictive method of the invention, it is therefore necessary to appropriately size the duration of said observation windows T.analysis so as to avoid: - excessive “combustion check” and / or “electrode check” notifications, in case excessively narrow observation time windows T.analysis are chosen (e.g. daily), or, conversely,

[0184] - too sporadic notifications of “combustion check” and / or “electrode check” as a result of observation windows T.analysis, for example of monthly or yearly duration.

[0185] It is clear that several variants of the predictive method described above are possible for the man skilled in the art, without departing from the novelty scopes of the inventive idea, as well as it is clear that in the practical embodiment of the invention the various components described above may be replaced with technically equivalent ones. For example, nothing prevents the observation window T. analysis instead of being of a fixed and default duration, from being of a “variable” amplitude during the monitoring period of the burner 10 of the heating system, and adaptable to specific behaviours thereof.

[0186] For such purpose, the predictive method of the invention may provide and implement self-learning logics of the behaviour and operation of said burner 10, for example of the actual number and / or frequency of ignitions in a “learning” period, preferably coinciding with one or more observation windows T. analysis. In fact, it is possible that, as a rule, the ignitions of a burner 10 of a heating system tend to repeat according to substantially similar ways over time, the uses and consumptions by a user being generally equally constant and usual.

[0187] According to this variant, the predictive method of the invention may therefore:

[0188] - reduce the initially fixed and default duration of said observation window T.analysis if a significant increase in the number of ignitions and use of the burner 10 is observed during at least one learning period,

[0189] - increase said duration of the observation window T.analysis in the event of a prolonged inactivity of the burner (e.g. during holiday or vacation periods), i.e. when there may be a significant, albeit temporary, reduction in the use of the burner 10 of the heating system (this “distancing” the risk of malfunctions and errors thereof, given the “less intense” period of use).

[0190] Finally, the method for predicting the efficiency state and operation of a heating system, so far described without any limiting intent with reference to the ignition step of a burner 10, may be extended with minimal adaptations within the reach of a person skilled in the art, also to the subsequent calibration or modulation steps thereof.

[0191] In such case, for the calculation of the ionisation index I0N.INDEX, the ionisation current read and measured in an instant of the calibration and / or modulation step may be taken into account, for example, suitably chosen and / or predefined, instead of considering exclusively an ionisation current at the safetytime t. safety (or, equivalently, the other maximum or minimum values of the ionisation current, mentioned above).

[0192] Furthermore, according to this possible executive variant, instead of being a function and expression of flame absence (or unstable and / or incorrectly formed flame), as provided up to now, the error index ERROR.INDEX may be calculated considering other operating characteristics of the flame F, for example, and without any limiting intent, the value of the flame “detachment” and / or “lift-off” during the calibration or modulation step. For such purpose, the heating system may be arranged with the aforementioned flame sensors (for example, optical, temperature, ultrasound sensors, or the like) capable, for example, of detecting the intensity of the flame generated so as to compare it with admissible threshold values.

[0193] It is clear that with the method for predicting the efficiency and operation state of a heating system of the invention the declared aims are therefore achieved; in particular, providing for a method:

[0194] - able to notify in advance the need for an intervention and / or maintenance relative to the risk that said heating system reaches a lockout condition, and inform about the possible causes of the malfunction and / or wear;

[0195] - which does not require the knowledge and acquisition of the entire time profile of a combustion parameter, for example of an ionisation current, during at least one operating step of the same heating system (e.g., the ignition one);

[0196] - which may be implemented without the need of resorting to specific test steps which provide for the temporary interruption of the normal operation of the same heating system and / or excessive and / or additional energy consumptions and polluting emissions;

[0197] - simple to implement in the electronics and / or components already supplied with the heating system;

[0198] - quick to be implemented and executed both for heating systems comprising combustion electronic control systems (CCS) and pneumatic.

Claims

CLAIMS Predictive method of the performance state of a heating system, in particular of a generator of thermal energy, obtainable from the combustion of a fuel, comprising at least:- a burner (10) housed inside a combustion chamber (11), --- at least a valve (12) for the delivery and dosing of said fuel and a related combustion air supply system (13, 14), so as to define a suitable air-fuel mixture to the burner (10),- at least an electrode (17) for:- the ignition of said air-fuel mixture and the formation of a flame (F) on said burner (10), and / or- detecting at least one combustion parameter, such as an ionisation current (lON.tk), representative of the state and of the characteristics of said flame (F),- a control unit (16) capable of managing at least the functions of one or more of the aforementioned components and implementing one or more steps of said predictive method, characterised in that it comprises at least: a) the measurement and / or acquisition of values of said ionisation current (lON.tk) at at least one time instant tkof an operating step of said burner (10), said measurement and / or acquisition being carried out and / or repeated one or more times within an observation time window (T.analysis) of said burner (10), b) the determination of at least:--- an “ionisation index” (ION. INDEX) extrapolated from said values of the ionisation current (lON.tk) acquired in said observation time window (T.analysis),- an “error index” (ERROR.INDEX) extrapolated from the combustion errors of said burner (10) detected in said observation time window (T.analysis),in which said steps a) and b) are repeated identically for each of the observation time windows (T.analysis) provided, the combination of said ionisation (ION.INDEX) and error (ERROR.INDEX) indices enabling to assign a “normal” (CL.N), “abnormal” (CL. A) or “critical” (CL.C) operating class to said heating system. Predictive method according to claim 1 , characterised in that said combustion errors for the determination of said error index (ERROR.INDEX) are detected and / or determined:- starting from said values of the ionisation current (lON.tk) acquired in said observation time window (T.analysis), and / or- by direct evaluation of characteristic parameters of the presence / absence of flame (F) at the burner (10). Predictive method according to claim 1 and / or 2, characterised in that said measurement and acquisition of at least one value of said ionisation current (lON.tk) is repeated for each of “i” ignitions of said burner (10) found within each of said observation time windows (T.analysis). Predictive method according to one or more previous claims, characterised in that said values of the ionisation current (ION.tk) consist of the values measured and acquired at a time instant tkprior to a safety time t. safety, such as the maximum (ION. peak) and / or minimum (ION. valley) value and / or the value at the ignition time (lON.ignition) or the like. Predictive method according to one or more previous claims 1 to 3, characterised in that said measured and acquired values of the ionisation current (lON.tk) consist of the flame value (lON.flame) calculated at a default safety time t. safety. Predictive method according to one or more previous claims, characterised in that it notifies a potential and / or already existing problem of said heating system if an “abnormal” or “critical” operationthereof has been found for a number N, M of consecutive observation time windows (T.analysis). Predictive method according to the previous claim, characterised in that it notifies a: - “combustion check”, if said “abnormal” operation was detected forN consecutive observation time windows (T.analysis), with N = 1, 2, ..., n, said notification concerning a potential and / or imminent malfunction state of the heating system attributable to combustion and / or installation and commissioning defects; - “electrode check”, if said “critical” operation was detected for M consecutive observation time windows (T.analysis), with M = 1, 2, ..., n, said notification concerning a malfunction state of the heating system attributable to the combustion unit (1) and related to an excessive oxidation and / or deformation and / or breakage of said at least an electrode (17) and / or to a breakage of said burner (10). Predictive method according to one or more previous claims, characterised in that from said values of the ionisation current (lON.tk) measured and acquired in a same observation time window (T.analysis), and for each of them, there are extrapolated at least: - one “ionisation index” (ION. INDEX) representative of cold ignitions of the burner (10),--- one “error index” (ERROR. INDEX) representative of combustion errors of the burner (10). Predictive method according to the previous claim, characterised in that said “ionisation index” (ION. INDEX) is a value determined: --- by filtering, from said values of the ionisation current (lON.tk), measured and acquired in an observation time window (T.analysis), the ones corresponding to:- cold ignitions of said burner (10) occurred after a cooling time t.cooling, and / or- internal temperatures of the heating system, in the proximity of or at the said burner (10), lower than a predetermined threshold value,--- by applying on said filtered and residual values of said ionisation current (lON.tk) different mathematical calculations comprising at least one operation which may consist of: calculating the maximum, calculating the minimum, calculating the average, calculating the trend or the slope of a curve interpolating said residual values, carrying out statistical processings.

10. Predictive method according to claim 8, characterised in that said “error index” (ERROR. INDEX) is determined by considering ignition errors corresponding to a failed and / or unstable activation of a flame (F) at the burner (10) in an observation time window (T.analysis), said ignition errors being able to cause blockages of said heating system.

11. Predictive method according to the previous claim, characterised in that said error index (ERROR.INDEX) is substantially representative of the flame absence frequency at the burner (10), said error index (ERROR.INDEX) being able to take an absolute value comprised between 0 and 1, where “0” indicates that no ignition errors occurred within an observation time window (T.analysis) of the burner (10), while “ 1” indicates a high frequency of ignition errors distributed in the same observation time window (T.analysis).

12. Predictive method according to one or more previous claims, characterised in that: - said “normal” operating class (CL.N) is substantially characterised at least:- by an error index (ERROR.INDEX) approximately or equal to 0, regardless of the value of the ionisation index (ION. INDEX), and / or - by low or very low values of the error index (ERROR.INDEX)and predominantly medium or high values of the ionisation index (ION.INDEX), said ionisation index (ION.INDEX) and error index (ERROR.INDEX) values being: ION.INDEX ≥ LIM.1 and ERROR.INDEX ≤ LIM.3;--- said “abnormal” operating class (CL. A) is substantially characterised by medium or high values for both the error index (ERROR.INDEX) and the ionisation index (ION.INDEX), said ionisation index (ION.INDEX) and error index (ERROR.INDEX) values being: ION.INDEX ≥ LIM.2 and ERROR.INDEX ≥ LIM.3a-- said “critical” operating class (CL.C) is substantially characterised by at least medium or high values of the error index (ERROR.INDEX) and by low values of the ionisation index (ION.INDEX), said ionisation index (ION.INDEX) and error index (ERROR.INDEX) values being: LIM. 1 ≤ ION.INDEX ≤ LIM.2 and ERROR.INDEX ≥ LIM.3b, where:- LIM.1 and LIM.2 consist of predefined limit thresholds, respectively lower, also called of flame detection, and higher, for said ionisation index (ION.INDEX),--- LIM.3, LIM.3a, LIM.3b consist of predefined limit thresholds for said error index (ERROR.INDEX). Predictive method according to one or more of the previous claims, characterised in that said observation time window (T.analysis) has a predetermined duration equal to one day, one week, one month, one year, or multiples or fractions thereof. Predictive method according to at least claim 1 and / or 2, characterised in that said measurement and acquisition of at least one value of said ionisation current (lON.tk) are performed during the calibration or modulation step of the burner (10).

15. Predictive method according to claim 14, characterised in that:--- said ionisation index (ION.INDEX) is a function of the values of the ionisation current (lON.tk) measured at an instant tkof said calibration or modulation step, - said error index (ERROR.INDEX) is a function of the operating characteristics of the flame (F), such as the flame (F) rise and / or detachment from the burner (10), in such case there being provided optical, temperature and / or ultrasound and / or other similar flame sensors adapted to detect said characteristics.

16. Heating system, in particular a thermal energy generator obtainable from the combustion of a fuel, comprising at least:- a burner (10) housed inside a combustion chamber (11),--- at least a valve (12) for the delivery and dosing of said fuel and a related combustion air supply system (13, 14), so as to define a suitable air-fuel mixture to be supplied to the burner (10),- at least an electrode (17) for:- the ignition of said air-fuel mixture and the formation of a flame (F) on said burner (10), and / or- detecting at least one combustion parameter, in particular an ionisation current (lON.tk), representative of the state and of the characteristics of said flame (F),- a control unit (16) capable of managing at least the usual functions of one or more of the aforementioned components, characterised in that said control unit (16) is capable of further implementing one or more steps of said predictive method of the performance state of said heating system according to one or more of claims 1 to 15.

17. Heating system according to claim 16, characterised in that said control unit (16) is the same control unit which subtends the general operation of said heating system and / or an ad hoc external control unitintegrated / integrable with said heating system and / or a remote processing logic that is implemented in the Cloud and / or in external mobile devices.

Images