Method of determining a decision indicator for selecting an optimal air filter and method of predictive maintenance of the filter

Abstract

The present invention is related to methods, particularly computer-implemented methods for selecting an optimal air filter for a given air treatment or air conditioning unit, by determining a decision indicator and wherein the optimal air filter makes the decision indicator minimal (or maximal). Additionally, the present disclosure is related to a method for predictive maintenance of a filter.

Description

METHOD OF DETERMINING A DECISION INDICATOR FOR SELECTING AN OPTIMAL AIR FILTER AND METHOD OF PREDICTIVE MAINTENANCE OF THE FILTERTechnical field

[0001] The present invention is related to methods, particularly computer- implemented methods for selecting an optimal air filter for a given air treatment or air conditioning unit, by determining a decision indicator and wherein the optimal air filter makes the decision indicator minimal (or maximal). Additionally, the present disclosure is related to a method for predictive maintenance of a filter.Background art

[0002] Air treatment units (ATUs), particularly for heating, ventilation and air conditioning (HVAC) applications, generally comprise one or more air filters to remove particulate matter from an air stream. The air is blown through the one or more air filters by a fan system. To do so, the fan consumes electrical energy to overcome a pressure drop of the air stream caused by the air filter. It has been estimated that about 70% of the operational costs of the ATU is due to the energy consumption of the fan, another 15% is due to the purchase cost of the air filter and the remainder being due to maintenance (filter replacement) and filter disposal costs.

[0003] Ventilation makes up a significant share of the energy consumption of modern buildings, and consequently of the operational costs. It is therefore of paramount importance to be able to reduce the operational expenses of the ATUs, particularly to reduce the energy consumption of the ATUs.

[0004] An air filter removes particles from an air stream by rejecting the particles at the upstream side of the air filter, where the particles (referred to as dust) accumulate on the media of the air filter. As dust is accumulating on the air filter, the pressure differential (pressure drop) across the air filter for maintaining a given flow rate of air increases, and so increases the power required by the fan. Air filter manufacturers specify for each air filter a dust holding capacity that is reached at an end of life of the air filter, when the air filter should be replaced, and associated end- of-life pressure drop. Different types of air filters have different dust holdingcapacities and different pressure drop characteristics making it difficult to select the optimal air filter from an energy efficiency and / or cost perspective.

[0005] Document US 11 ,719,454 discloses a method for determining a filter selection for an air filtration system having a current air filter, comprising the steps of: i. entering filtration system information into a processor having access to pressure drop versus dust holding capacity curves for a plurality of different air filters, wherein the filtration system information includes user experience information including volume of air processed with the current air filter in the air filtration system and pressure drop with the current air filter after processing the volume of air; ii. determining a cost factor of operating the air filtration system with the current filter for processing the volume of air; determining a theoretical quantity of dust held by the current filter at the pressure drop using a pressure drop versus dust holding capacity curve for the current air filter and the pressure drop exhibited by the current air filter after processing the volume of air; iii. selecting a proposed air filter from the plurality of different air filters and different from the current air filter; determining an estimated volume of air that can be processed with the proposed air filter to reach the pressure drop, this determining step being conducted using the theoretical quantity of dust and a pressure drop versus dust holding capacity curve for the proposed air filter; iv. determining an estimated cost factor of operating the air filtration system with the proposed filter for processing the volume of air; and v. presenting a comparison of the cost factor and the estimated cost factor on a display communicated with the processor.

[0006] The method described herein above relies on user experience information based on a particular filter system provided with a first filter (i.e. a current filter) the measurement of a volume of air processed, and on a choice of a proposed filter to establish a comparison of costs of processing a same volume of air through a same filter system when using the first filter or the proposed filter. A proposed filtercan be selected by the user from a library of different filters with which the system is communicated. The total available base of filters can itself be filtered to only allow selection by the user of filters that will fit the air filtration system in question and are adequate for any particular issues at that location. A user of a filtration system may lack of experience information and of particular issues of the location, especially for a new filtration system, so that selection of an appropriate filter requires more guidance.

[0007] Document EP2105677 describes a method for selecting a filter wherein in a first step, S1 , a pollution level is measured of the air to be received by the filter system and this pollution level is the first pollution level. In a subsequent step, S2, a filter parameter is determined based on a combination of the measured pollution level and a second desired pollution level. The filter parameter can be a filter class of the EN779:2002 classification system or an amount reduction of PM 10 performed by the filter in a percentage value. On to the next step, S3, a filter of the determined filter class is installed in the filter system. In the next step, S4, a pollution level of the air provided by the filter system in which the selected filter is arranged, is measured. In step S5 the measured pollution level is compared with the second pollution level. Then, in step 6, it is questioned if the measured pollution level is in a predetermined acceptable error range. If the answer is yes, the method of selection ends, otherwise another filter is selected and steps S4 to S6 are reiterated. of the invention

[0008] There is hence a need in the art to provide a method of selecting an optimal air filter of an air treatment unit allowing to optimize energy efficiency and / or reduce costs, particularly operational costs of the ATUs including the air filter. In view of environmental concerns, especially to reduce carbon footprint, there is a need in the art to provide methods of selecting an optimal air filter of an air treatment unit allowing to reduce a carbon footprint of the ATU.

[0009] There is also a need to adjust the moment when a filter of an ATU must be changed, in function of the real conditions of filtration.

[0010] According to a first aspect of the present disclosure, there is therefore provided a method of determining a decision indicator for selecting an optimal air filter of an air treatment unit, as set out in the appended claims. The method isparticularly a computer-implemented method, allowing to select an optimal air filter for the air treatment unit based on a decision or selection criterion which is based on an electrical energy consumption of the air treatment unit due to the air filter.

[0011] One advantage of methods of the present disclosure is that the electrical energy consumption, and hence a decision or selection criterion to select an optimal air filter for a given air treatment unit, is determined based on an air quality indicator that is representative of a particulate matter concentration of the dust laden air that is fed to the air filter at an installation location of the air treatment unit. The air quality indicator is advantageously obtained by querying an air quality database based on a location indicator of the air treatment unit. By so doing, a more accurate estimation of the service life of the filter and its clogging state is obtained, enabling a more accurate determination of the expenditures relating to selecting a particular filter. Alternatively, the air quality indicator can be obtained by measurement of air quality (e.g., particle matter concentration) of the dust laden air fed to the air treatment unit, e.g. at an inlet of, or in proximity of the air treatment unit. In addition, methods of the present disclosure further enable to optimize operating conditions of the air treatment unit and / or of the air filter in terms of reducing electrical energy consumption and / or optimizing a selection or decision criterion.

[0012] According to a second aspect of the present disclosure, there is provided a data processing system, as set out in the appended claims.

[0013] According to a third aspect of the present disclosure, there is provided a computer program product, as set out in the appended claims

[0014] According to a fourth aspect of the present disclosure, there is provided a method for predictive maintenance of a filter used in an air treatment unit.Brief description of the drawings

[0015] Aspects of the invention will now be described in more detail with reference to the appended drawings, wherein same reference numerals illustrate same features and wherein:

[0016] Figure 1 represents a diagram of an air treatment unit according to the present disclosure;

[0017] Figure 2 represents a top view of a pocket air filter that can be utilized in the present disclosure;

[0018] Figure 3 represents a flow diagram of an embodiment of a method according to the present disclosure;

[0019] Figure 4 represents a flow diagram of another embodiment of a method according to the present disclosure;

[0020] Figure 5 represents a diagram of a data processing system according to the present disclosure.Detailed description of the invention

[0021] Referring to Fig. 1 , an air treatment unit (ATU) 10 comprises a duct 11 through which air is made to flow in a flow direction indicated by arrow 100 from an inlet 101 of the duct to an outlet 102 of the duct 11 . The air flow can be generated by a flow fan 13 or other air flow generating system, such as an air pump. A filter system 12 is provided in duct 11. Filter system 12 can comprise one or more air filters 122, each mounted in a filter frame 121 advantageously providing an air-tight sealing between the air filters 122 and the duct 11. The air flow 100 through the duct11 passes integrally through the air filter 122. The filter system 12 can be arranged upstream or downstream with respect of the flow fan 13, in duct 11.

[0022] ATU 10 can comprise a pressure sensor 141 upstream of the filter system 12 to determine a (static and dynamic) pressure Pj of the air stream at an inlet of the filter system and / or a pressure sensor 142 downstream of the filter system12 to determine a (static and dynamic) pressure Po of the air stream at an outlet of the filter system. Each pressure sensor 141 , 142 can be coupled to a control unit 15 which may monitor a performance of the filter system 12. Control unit 15 can be operably coupled to the flow fan 13, e.g. to control or adjust a rotation speed and / or an air flow rate through the duct. To this end, control unit 15 can comprise a data input port 151 to receive additional measurement data, e.g. from other sensors, and / or control commands, e.g. from a user interface, enabling to control the air flow through the duct 11 .

[0023] ATU 10 can comprise a particle concentration sensor 143 configured to measure a particulate matter concentration (e.g. PM™ and / or PM2.5) or dust loadingin the air stream at an inlet of the filter system 12 and / or a particle concentration sensor 144 configured to measure a particulate matter concentration or dust loading in the air stream at the outlet of the filter system. Each particle concentration sensor 143, 144 can be coupled to control unit 15.

[0024] The ATU 10 can form part of a larger air flow system, such as a system for heating, ventilation and air conditioning (HVAC). Particularly, ATU 10 can be the intake of a HVAC system, configured to take outdoor air through inlet 101 , passing it through filter system 12, possibly conditioned by additional conditioning systems which may be provided in the duct 11 , and supplying it to the HVAC system via outlet 102. Alternatively, ATU 10 can form part of an air recirculation system, in which exhaust air from an indoor space or building is drawn through inlet 101 to be filtered through filter system 12, possibly reconditioned by additional conditioning systems which may be provided in the duct 11 , and reinjected in the indoor space of building through outlet 102. Yet alternatively, ATU 10 can form part of an exhaust system of an industrial plant, in which exhaust gasses are collected through inlet 101 , passed through filter system 12, e.g. to collect dust and / or contaminants, possibly through other gas / air cleaning systems provided in the duct 11 and released into the atmosphere through outlet 102.

[0025] Referring to Fig. 2, the filter system 12 can comprise one or a plurality of air filters 122, each mounted in a filter frame 121. Multiple air filters 122 can be arranged in an array or matrix fashion to span an entire cross sectional area of duct 11 , as known in the art. The air filters can be pocket filters, in which the filter media is shaped into pockets 1221 , 1222, 1223 to increase the available surface area for filtering. Alternatively, the air filter can be a pleated filter, a planar filter, or any other suitable filter as known in the art for filtering air or gas, in particular combination filters such as an RPV filter comprising a rigid frame defining end caps for the pockets and a possibly pleated filter media defining walls of the pockets.

[0026] The air filters 122 are specifically configured to act as particle filters, such as to reject particulate matter (PM) from the air stream passing through the filter media. The air filters can be made of any suitable filter media as known in the art, such as paper, woven or nonwoven materials, including synthetic (polymeric) materials such as polypropylene or inorganic materials such as glass fibres. The air filters have an upstream side 123 and a downstream side 124, with the stream of airconfigured to pass from the upstream side 123 to the downstream side 124 through the air filter 122. As air is passing through the air filter 122, particulate matter is held up by the filter media of the air filter 122 and removed from the air supplied to the downstream side 124. This particulate matter accumulates on and in the filter media. As more particulate matter accumulates, the filter pores get increasingly clogged, reducing the size of the filter pores, leading to enhanced particulate matter rejection but increased flow resistance through the air filter 122.

[0027] Typically, a relationship between the amount of accumulated particulate matter, and the flow resistance through the filter is non-linear. The flow resistance can be expressed in terms of a pressure differential AP = Pj -Pobetween the upstream side 123 and the downstream side 124 of the air filter. Since the pressure differential is proportional to flow resistance, the pressure differential will relate to the amount of accumulated particulate matter, typically in non-linear fashion.

[0028] Air filters are currently categorized according to the ISO 16890 - 1 :2016 standard. The standard establishes a filter efficiency classification system based on particulate matter size that the filter is able to reject. Particulate matter in an air stream or aerosol is expressed as PMXwith x referring to particles having an optical diameter smaller than or equal to x pm. In some examples, particles with optical diameter smaller than 0.3 pm are not taken into account. Four efficiency classes are established in increasing coarseness level: ePMi, ePIVk.s, ePM and ISO Coarse. An air filter belongs to the efficiency class with lowest PM size for which the filtration efficiency (e.g., the rejection ratio) is higher than or equal to 50%, or to the ISO Coarse class otherwise, e.g. if the filter removes less than 50% of PM™. The filter efficiency is determined based on laboratory experiments involving a synthetic test dust specified in EN ISO 15957 as L2 (AC Fine). In practical applications, the dust loading of the air stream to which the filter is exposed may deviate significantly from the synthetic test dust according to the standard, resulting in a different filter efficiency performance.

[0029] The energy consumption of the flow fan 13 in ATU 10 can be evaluated based on the volume flow rate displaced by the fan, the fan efficiency coefficient, the operating time, and the pressure differential AP. Typically, the efficiency of the fan is a function of the fan speed. The actual fan efficiency coefficient can also strongly depend on the design and the layout of ATU. Typical values of the fan efficiencycoefficient vary between 0.80 and 0.25. Eurovent Industry Recommendation 4 / 24 - 2023 Energy consumption evaluation of air filters for general ventilation purposes, https: / / eurovent.eu / ?q=content / eurovent-424-2023-energy-consumption- evaluation-air-filters-general-ventilation-purposes, discloses a generic methodology for estimating energy consumption of a single air filter according to the following equation:wherein W represents total energy consumption in kWh, represents the air volume flow rate in m3 / s, t represents the operating time in hours, q represents fan efficiency coefficient, and Ap represents the average pressure drop of the filter in Pa.

[0030] An air filter is said to be at its end-of-service-life either when the particulate matter accumulated on and in the filter media of the air filter reaches a predetermined threshold, or when the air filter has been utilized for a predetermined lifetime. Once the predetermined threshold or the predetermined lifetime passed, the filter should be replaced, involving disposal costs, replacement costs and costs for purchasing a new filter. Eurovent Industry Recommendation 4 / 24 - 2023 hence differentiates between two filter replacement criteria: (i) a time-based method in which the air filter is changed according to fixed time schedule, and (ii) a condition based method in which the air filter is changed when having reached a final pressure drop (i.e. , pressure differential), and consequently a final (or end-of-service-life) dust loading, also referred to as dust holding capacity.

[0031] In the time-based method (i), the operating time t in Eq. 1 is known and the average pressure drop Ap is calculated as an integral average pressure drop from a laboratory test method in which the instantaneous pressure drop is measured for incremental dust (mass) loadings to obtain different data points (m^Ap,) in a pressure drop (Ap) versus dust loading (m) curve. Specifically, standard ISO L2 dust (AC Fine) (ISO 15957) is fed as dust according to the procedure set out in ISO 16890-3 in incremental amounts to reach at least an amount of dust loading Mxthat is dependent on the efficiency class of the filter and on the flow rate (^^-.200g for ePMi;for ePM ). The dust loading amounts are estimates of real-life dust concentrations: ePMi = 10 pg / m3; ePIVh.s « 12 pg / m3;ePM » 20 pg / m3. The average pressure drop Ap is calculated according to a weighted average of the average pressure drops Ap of each measured dust loading interval Am; with the last interval cut off at dust loading Mx.wherein Mx== i^ rrtt - mi-1is the total dust mass fed in which m0= 0 g; and mn= Mx,Wi = 0.5 (Ap; T Api- ; and Ap^ = 0.5where mnis the first measurement point larger than or equal to Mx.

[0032] In the condition-based method (ii), the average pressure drop Ap in Eq. 1 is calculated based on the initial pressure drop Ap0of the filter at no dust loading, on the final pressure drop that is selected for its end-of-service-life and on the shape of the pressure drop curve. According to Eurovent Industry Recommendation 4 / 24 - 2023, an estimation of the average pressure drop Ap can be expressed as:The final pressure drop ^Pfinaimay be selected as a multiple of the initial pressure drop or as the initial pressure drop increased by a predetermined pressure differential to simplify calculations.

[0033] According to Eurovent Industry Recommendation 4 / 24 -2023, the above calculations of Eq. 1 - Eq. 3 apply to filters with face dimensions of standard size 592 mm x 592 mm. To determine energy consumption of a filter with different dimensions, the test must first be carried out with a standard sized filter of the same family, and the energy consumption of the non-standard filter determined by assuming that the ratios of energy consumption W (Eq. 1) to air volume flow rate qvare identical for the standard and non-standard filter.

[0034] Eurovent Industry Recommendation 4 / 23 - 2022 Selection of EN ISO 16890 rated air filter classes for general ventilation applications, https: / / eurovent.eu / ?q=content / eurovent-423-2022-selection-en-iso-16890-rated-air-filter-classes-fourth-edition-english provides guidelines on selecting an appropriate filter class by matching the minimum filtration efficiency of the filter with both the outdoor air and the supply air quality. For the supply air, thresholds for PM concentrations established by the World Health Organization (WHO) are considered. Currently recommended annual mean limits (WHO guidelines 2021) to be observed are an annual mean for PM2.5 < 5 pg / m3and annual mean for PM™ < 15 pg / m3. WHO guidelines do not currently set a limit on PM1 concentrations since no threshold was identified below which no damage to health is observed. Three categories of outdoor air (ODA) quality and five categories of supply air (SUP) quality are defined.

[0035] The ODA categories are:ODA 1 : Outdoor air, which may be only temporarily dusty, corresponding with outdoor air meeting the annual mean limits established by WHO.ODA 2: Outdoor air with high concentrations of particulate matter, corresponding with outdoor air where PM concentrations exceed the annual mean limits established by WHO by a factor up to 1 .5, i.e., an annual mean for PM2.5 7.5 pg / m3and annual mean for PM™ s 22.5 pg / m3.ODA3: Outdoor air with very high concentrations of particulate matter, corresponding with outdoor air where PM concentrations exceed the annual mean limits established by WHO by more than a factor 1.5, i.e., an annual mean for PM2.5 > 7.5 pg / m3and annual mean for PM™ > 22.5 pg / m3.Determination of ODA category depends on the local outdoor air quality, which may be assessed based on the air pollution database published by WHO indicating air pollution monitoring results of multiple locations worldwide and regularly updated.

[0036] The SUP categories are:SUP 1 : Supply air with particulate matter concentrations fulfilling the annual mean values established by WHO multiplied by a factor 0.25, i.e., an annual mean for PM2.5< 1.25 pg / m3and annual mean for PM™ s 3.75 pg / m3. The SUP 1 category is recommended for applications with high hygienic demands, such as hospitals.SUP 2: Supply air with particulate matter concentrations fulfilling the annual mean values established by WHO multiplied by a factor 0.5, i.e., an annual mean for PM2.5< 2.5 pg / m3and annual mean for PM™ s 7.5 pg / m3. The SUP 2 category isrecommended for rooms for permanent occupation and / or with medium hygienic demands, such as offices, residential buildings, theatres, cinemas, spaces for food and beverage production.SUP 3: Supply air with particulate matter concentrations fulfilling the annual mean values established by WHO multiplied by a factor 0.75, i.e., an annual mean for PM2.5< 3.75 pg / m3and annual mean for PM 10 < 11.25 pg / m3. The SUP 3 category is recommended for rooms with temporary occupation and / or basic hygienic demands, such as shopping centres and server, storage or copier rooms.SUP 4: Supply air with particulate matter concentrations fulfilling the annual mean values established by WHO multiplied by a factor 1 , i.e., an annual mean for PM2.5 5 pg / m3and annual mean for PM™ s 15 pg / m3. The SUP 4 category may be utilized for rooms with short-term occupation and / or without hygienic demands, such as restrooms or general industrial manufacturing halls.SUP 5: Supply air with particulate matter concentrations fulfilling the annual mean values established by WHO multiplied by a factor 1. 5, i.e., an annual mean for PM2.5< 7.5 pg / m3and annual mean for PM™ s 22.5 pg / m3. The SUP 5 category may be utilized for rooms without occupation, such as underground car parks or garbage rooms and / or for production areas of heavy industries.

[0037] Based on the ODA category of the location or environment in which the air filter is placed and the desired SUP category, Eurovent Industry Recommendation 4 / 23 - 2022 sets out recommended minimum filtration efficiencies to aid users in selecting an appropriate air filter, e.g. for ODA 2 and SUP 2 category, an ISO ePMi 70% air filter is recommended, whereas for ODA 2 and SUP 3 category, an ISO ePM2.5 70% air filter is recommended. In practical applications, filter manufacturers or facility operators can determine the class of filters to utilize for a given ATU installation based on Eurovent Industry Recommendation 4 / 23 - 2022, and determine the most energy-efficient filter within this class based on Eurovent Industry Recommendation 4 / 24 - 2023 and the filter specifications (e.g., pressure drop versus dust loading curve) provided by the filter manufacturer. One drawback in the above approach is that the calculations set out in Eurovent Industry Recommendation 4 / 24 - 2023, e.g., Eq. 1 - Eq. 3, do not fully account for the actual dust loading at the location of the ATU installation. In addition, additional costs and expenditures can apply relating to filter purchase, replacement and disposal whichmay be significant when attempting at selecting the most cost-effective filter for the given ATU.

[0038] Referring to Fig. 3, a method 200 of determining an optimal filter for an ATU selected out of a plurality of filters, such as a list of available filters for the ATU, comprises an operation 201 in which operating data of the ATU is received. The operating data is representative of operating conditions of the ATU and comprises a volumetric air flow rate indicator 211 of the ATU, an energy efficiency indicator 212 of the ATU and possibly a supply air quality indicator 213 and / or a location indicator 214 of the ATU. The volumetric air flow rate indicator 211 refers to data enabling to determine a representative volumetric air flow rate qvfor a single air filter and can comprise a total volumetric air flow rate supplied by the ATU and possibly a number of air filters disposed in parallel in the duct of the ATU. Operation 201 can comprise determining the volumetric air flow rate qvfor a single air filter based on the volumetric flow rate indicator 211 , e.g. through dividing the total volumetric air flow rate by the number of air filters. The volumetric air flow rate indicator 211 can refer to, or represent, a mean volumetric flow rate in case the ATU is operated at variable volumetric air flow rates, or can enable to determine such mean volumetric flow rate. It will be appreciated that the volumetric air flow rate qvfor a single air filter refers to a mean volumetric flow rate over a given period, e.g. over the filter service life.

[0039] The energy efficiency indicator 212 refers to data enabling to determine the electric power P drawn by the ATU based on the volumetric air flow rate of the ATU and the pressure differential across the air filter, e.g. based on the equation:which can easily be derived from Eq. 1. Advantageously, the energy efficiency indicator 212 refers to the efficiency coefficient^ of the flow fan of the ATU.

[0040] The supply air quality indicator 213 comprises data representative of the air quality which the ATU supplies at its outlet, i.e. of the air quality the air filter needs to provide. In some examples, the supply air quality can be provided as one of the supply air categories SUP 1 , SUP 2, SUP 3. Additional or alternative supply air categories can be contemplated as desired. The supply air quality indicator 213 advantageously enables to select appropriate air filters or air filter categories for the ATU allowing to meet the predetermined supply air quality as expressed by thesupply air quality indicator 213. Alternatively, or in addition, the supply air quality indicator 213 is representative of one or more classes, types or categories of air filters which are specified for installation in the ATU.

[0041] The location indicator 214 comprises data representative of a location in which the ATU is installed. This may refer to a geographical location, e.g. supplied as position coordinates, or as a place indication including one or more of: a name of a municipality, a postal code, and a country and / or region code. Alternatively, or in addition, the location indicator 214 may refer to data indicating an environment in which the ATU is installed, e.g. a rural, suburban or urban environment.

[0042] In operation 202, an inlet air quality indicator 215 is received. The inlet air quality comprises data representative of a pollution, such as a particulate matter pollution or of a dust loading, of the air that is supplied to the ATU and which hence is fed to the (upstream side of the) air filter. This may be outdoor air, depleted indoor air, or exhaust air or gas, e.g. from an industrial plant. Particularly, the inlet air quality indicator comprises data indicative of particulate matter concentration, e.g. PM™ concentration and / or PM2.5 concentration, or any other data indicative of a dust loading of the air that is fed to the air filter. Operation 202 can comprise determining a particulate matter concentration, or a dust loading of the inlet air PMjnof the ATU based on the inlet air quality indicator. PMjnadvantageously enables to determine a dust loading of the air filter, and PMjncan correspond with the inlet air quality indicator 215, or can be derived therefrom.

[0043] The inlet air quality indicator 215 can be obtained by querying one or more databases, which may be public databases with air pollution data, such as the WHO Ambient Air Quality Database (www.who.int), based on a location of the ATU, particularly based on the location indicator 214, specifically if outdoor air is supplied to the ATU. If the location indicator 214 comprises a geographical location, inlet air quality data may be obtained by querying an air quality database based on the geographical location, e.g. country, region and / or city. Alternatively, if the location indicator comprises data indicating an environment in which the ATU is installed, the inlet air quality indicator 215 can be determined from representative air quality parameters (e.g. PM concentration) of the respective environment, e.g. by determining an average air quality parameter from measurements at different locations corresponding with the given environment. By way of example, the locationindicator 214 may indicate that the ATU is installed in Belgium in a suburban environment, and the inlet air quality indicator 215 may provide air quality measurements from various Belgian suburban environments. Yet alternatively, air quality measurements, such as PM concentration data, of the inlet air at the ATU instalment may be available, e.g. due to PM sensors comprised in the ATU, and the inlet air quality indicator 215 can refer to data obtained from the measurements. Operation 202 may determine a mean inlet air quality from these air quality measurements as a representative inlet air quality for the ATU (PMjn). One advantage of operation 202 and / or of the inlet air quality indicator 215, is that a more accurate dust loading of the air fed to the air filter is obtained, which allows to determine an end-of-service-life of the air filter, and hence energy consumption of the air filter, more accurately.

[0044] Alternatively, in particular when the ATU 10 is configured to recirculate or recycle indoor air, the inlet air quality indicator 215 can be obtained by measurement of particle concentration of the air stream fed to the inlet of the filter system, e.g. through the particle concentration sensor 143 of the ATU 10 and / or through any air pollution or particle concentration sensor installed in an indoor environment and configured to sense an air quality representative of the air quality fed to the inlet of the filter system.

[0045] In operation 203, one or more operating attributes 216 of each of the (selected) plurality of air filters are provided. The operating attributes 216 advantageously comprise flow resistance indicators associated with the respective air filter, such as data representative of flow resistance (e.g., pressure differential) as a function of a dust loading of the respective air filter, e.g. m pt) data points. The operating attributes 216 advantageously enable to determine an average pressure drop Ap (or average flow resistance) of the air filter over its service life, e.g. for use in Eq. 1 or Eq. 4. The operating attributes 216 can comprise a filtration efficiency of the respective air filter, e.g. according to ISO 16890, such as ePMi, ePM2.5, or ePM .

[0046] Optionally, in operation 203, one or more cost attributes 217 are provided in operation 203. The cost attributes 217 can comprise one or more of: a purchase cost of the respective filter, a disposal cost of the respective filter and a replacement cost of the respective filter.

[0047] The one or more operating attributes 216 and possibly the one or more cost attributes 217 may be retrieved by querying one or more air filter databases. Such databases may comprise data on the operating attributes and the cost attributes of a plurality of air filters associated with one or multiple filter types, categories or classes. In operation 203, the one or more operating attributes 216 and possibly the one or more cost attributes 217 may be provided for a plurality of filters based on the supply air quality indicator 213. Advantageously, the supply air quality indicator 213 and possibly the inlet air quality indicator 215 is utilized in operation 203 to select suitable air filters from one or more air filter lists or filter databases. Advantageously, air filters are selected, e.g. from a list of air filters, that meet requirements for supply air specified by the supply air quality indicator 213 and / or the inlet air quality indicator 215, e.g. in terms of ISO filter efficiency such as described in Eurovent Industry Recommendation 4 / 23 - 2022. The selected plurality of air filters will be retained for the further operations of method 200. It is possible to further refine the plurality of air filters based on an appropriate decision logic, e.g. filter type or manufacturer. The one or more operating attributes 216 and possibly the one or more cost attributes 217 can be retrieved or provided for the selected air filters only. It will be appreciated that the filter lists, or filter databases from which suitable air filters are selected, and even the selected plurality of air filters can comprise filters of different type. One advantageous benefit of methods of the present disclosure is that filters of different types can be compared in terms of energy consumption or cost without difficulty and based on an identical set of filter attributes.

[0048] In operation 204, an end-of-service-life indicator EOL is determined, particularly for each filter selected in operation 203. The EOL indicator can refer to a final pressure differential (final pressure drop) or to a final dust loading (dust holding capacity) of the air filter. The EOL indicator can be determined based on the filter replacement strategy, such as time-based or condition-based, and the EOL indicator can comprise data representative of the filter replacement strategy. For a time-based replacement strategy, the EOL indicator can comprise a total operating time period of the respective filter, e.g. one year if the filter is replaced after one year of operation or installation. For a condition-based replacement strategy, the EOL indicator can comprise a maximal dust loading mass Mewhich may further be based on one or more of the operating attributes 216, e.g. based on the filtration efficiency of therespective filter, such as outlined in Table 1 of Eurovent Industry Recommendation 4 / 24 - 2023. It will be appreciated that filter replacement strategy may vary for different filters. For at least some filters, e.g. selected in operation 203, operation 204 can determine both a time-based EOL indicator and a condition-based EOL indicator.

[0049] It will be appreciated that the EOL indicator can be provided as a predetermined data input, e.g. as part of the operating attributes 216 of the respective filters, or part of operating data of the ATU. The EOL indicator may e.g. be predetermined by a filter manufacturer, by an operator of the ATU, or by legislation or standard.

[0050] In operation 205, an electrical energy consumption W of the ATU due to the respective filter is determined. The electrical energy consumption W can be determined for a service life of the respective filter TF, based on the EOL indicator, or for any other predetermined (fixed) operational time period of the ATU TA, e.g. a service life of the ATU, or other suitable (fixed) evaluation period. Such a predetermined operational time period can be provided as a time indicator 218 in operation 205.

[0051] The service life of the filter TF is determined based on the EOL indicator (operation 204). If the EOL indicator refers to a time-based replacement strategy, determination of T from the EOL indicator is straightforward. If the EOL indicator refers to a condition-based replacement strategy, e.g. the EOL indicator refers to a maximal dust loading, the service life T is advantageously determined based on the inlet air quality indicator 215 (via the dust loading PMjnof the air fed to the respective filter - operation 202) and on the volumetric air flow rate qvof the respective filter (operation 201), e.g. through the formula:wherein Meis the final dust loading of the filter (at end-of-service-life) [g], PMjnis the particulate matter concentration of the air fed to the air filter [pg / m3] and the volumetric air flow rate qvis expressed in m3 / s and T in hours. When the volumetric air flow rate qvis expressed in m3 / h, the factor 3600 at the denominator is absent in Eq. 5. If a final pressure drop instead of a final dust loading is indicated as EOLindicator, the final dust loading can be calculated based on the (mi, A.pi') data points of the respective filter.

[0052] The obtained value of TF for either filter replacement strategy can be utilized as parameter t in Eq. 1 if the energy consumption for a service life of the air filter is desired. The average pressure drop Ap is advantageously determined based on the EOL indicator, e.g. by considering the maximal dust loading mass Me= Mxin Eq. 2 for condition-based filter replacement strategy. For time-based filter replacement strategy, TF is specified by the EOL indicator and Meis determined based on the EOL indicator, the inlet air quality indicator 215 and the volumetric air flow rate qvof the respective filter, e.g. through the relationship of Eq. 5. Hence, both in time-based and condition-based strategies, the air quality indicator 215 (possibly in conjunction with the location indicator 214) advantageously enables to obtain a more accurate estimation of filter service life or final dust loading, and hence of the energy consumption related to the filter. The electrical energy consumption is advantageously determined for each filter selected or determined in operation 203.

[0053] In operation 206, to compare different filters, which may have different service lives TF and / or final dust loading masses Me, in terms of energy efficiency, it is more convenient to determine energy consumption WTover a (fixed) predetermined operational time period TA of the ATU, e.g. one year or a plurality of years or an entire service life of the ATU. Advantageously, TA S TF in respect of all the filters that are compared. Typically, TA will not correspond with integer multiples of TF. Considering:TA / TF = R with R being a real number,R = N + r with N being a natural number and 0 < r < 1 , the total energy consumption WTcan be expressed as:WT= N ■ W + WF(Eq. 6), where W represents energy consumption in respect of a single filter for its service life and WFrepresents a fractional energy consumption of the last filter that will be installed in period TA and that will not be utilized over its service life. WF= 0 if r = 0. To determine WFwhen r is nonzero, Eq. 1 can be utilized, corrected for the actual dust loading of the filter over the time period t = r.Tp. Hence, the calculations ofoperation 205 can be performed with TF replaced by r.Tp to determine a corresponding dust loading Meand average pressure drop Ap . WFcan be obtained through Eq. 1 with the adapted values t and Ap.

[0054] By so doing, a more accurate comparison of different possible filters, based on actual energy consumption and a more accurately determined service life or final dust loading, is obtained.

[0055] In a possible further operation 207, an operating cost Cu of the ATU is determined, based on the total energy consumption WTand associated energy cost attribute 219. The energy cost attribute 219 comprises data allowing to determine an economical or environmental expenditure of electrical energy, e.g. a price per kWh or a carbon footprint per kWh. The operating cost Cu is advantageously determined based on the location indicator 214, e.g. to retrieve the energy cost attribute 219 associated with a geographical location (country, region, municipality) determined based on the location indicator 214. The energy cost attribute 219 may be retrieved by querying a database comprising geographical locations and associated electrical energy costs. If the energy cost determined from energy cost attribute is pe, the operating cost Cu can be determined as:The operating cost Cu can hence refer to an economical cost, and / or to an environmental cost, e.g. a carbon footprint, determined based on the total energy consumption WT.

[0056] In operation 207, a total cost of ownership TCO of the ATU can further be determined. The TCO is determined based on the operating cost Cu and on the filter cost attribute(s) 217. Particularly, any one of: a purchase cost of the respective filter, a disposal cost of the respective filter and a replacement cost of the respective filter can be taken into account to determine TCO, in which these terms should be multiplied by the number of filters that are purchased, replaced and / or disposed of over the time period TA, e.g. based on R. The filter cost attribute(s) 217 can be retrieved from a database, e.g. by querying based on the location indicator 214. The filter cost attribute(s) 217 can refer to an economical cost, and / or to an environmental cost, e.g. a carbon footprint.

[0057] It will be appreciated that any one of the (total) electrical energy consumption W, WT, operating cost Cu, the total cost of ownership TCO, or any other suitable indicator based on any of the above parameters can be utilized in methods according to the present disclosure as a decision indicator to select an optimal air filter.

[0058] Referring to Fig. 4, a method 300 differs from method 200 of Fig. 3 in the fact that one or more iterations are performed to further optimize a decision indicator, referring to any one of the (total) electrical energy consumption, operating cost or TCO. The optimization advantageously comprises adjusting the filter service life TF such as in view of optimizing the value of r and hence the total energy consumption WTin Eq. (6). This can be performed through a test operation 301 to determine whether an optimum is reached in the decision indicator and iteration loop 302 that adjusts the EOL indicator (service life TF and / or final dust loading Me) according to a predetermined logic to adjust or further optimize the total energy consumption WT. In some examples, the test operation 301 , may determine to adjusting R to the closest higher ranked integer number, e.g. R = N +1 if r is nonzero. Any other suitable optimization logic, e.g. a least squares optimization, can be contemplated for test operation 301 to find a minimum WTand / or TCO for each of the air filters. An optimal filter 302 is obtained as positive outcome of operation 301 relating to an extremum of a decision or selection criterion (e.g. minimal TCO or electrical energy consumption).

[0059] Referring to Fig. 5, a data processing system 40 comprises a processing module 41 , a memory module 42 and a user interface 43 coupled via a data bus interface 47. The data processing system 40 may be coupled to one or more database systems 44, 45, which may be external database systems, via a data communication connection 46. The data processing system 40 can further be connected to the control unit 15 of the ATU 10 (not shown) via a suitable data communication connection (wired or wireless), e.g. to receive measurement data of one or more pressure sensors 141 , 142 and / or one or more particle concentration sensors 143, 144. Processing module 41 is configured to carry out instructions that perform methods of the present disclosure, particularly methods 200 and / or 300 as set out above. To this end, the memory module 42, which comprises permanent and / or volatile memory, can store these instructions as a computer program product.Memory module 42 can further store data representative of one or more of the indicators, e.g. indicator 219, while other indicators, e.g. indicator 215, 216, 217 may be fetched from the database systems 44, 45 via data communication connection 46. Memory module 42 can (temporarily) store data retrieved from database systems 44 and / or 45. Yet other indicators, specifically the indicators 211-214 and 218 relating to operating data of the ATU, may be provided via a user interface 43. Processing module 41 can be configured to transmit an outcome of the method, e.g. as a total energy consumption WTand / or TCO to user interface 43 enabling a user to select the optimal filter for the given ATU and operating conditions. Alternatively, processing module 41 can be configured to determine the optimal filter for the given ATU, as a result of applying methods of the present disclosure, and possibly indicate the optimal filter that is determined via the user interface 43.

[0060] In another aspect, a method for predictive maintenance of a filter used in an air treatment unit is provided. Preferably, an optimal filter for an air treatment unit is selected according to the method described herein above. Preferably, the optimal filter is a filter being the most efficient in term of energy consumption and total cost of ownership, and which is has the efficiency for providing a desired level of supply air quality (SUP) on a level of outdoor air quality (ODA) and.

[0061] During the time of utilization of the selected filter in the ATU, conditions of outdoor air quality may be subject to variations, other unexpected parameters may have an influence on the filtration efficiency, so that the end-of-service-life indicator determined may not reflect the optimum time conditions of use of the selected filter. In function of its conditions of utilization, when considering the end-of-service-life indicator for replacing the selected filter, there is a possibility that the selected filter is replaced too early while being still operational for providing the desired supply air quality, or inversely that the replacement of the filter is planned at a moment too late to allow a full efficiency of the filter over its full time of use in the ATU.

[0062] The method for predictive maintenance aims to solve this problem by the steps of : i. Establishing or obtaining for a selected filter, a first relation between the pressure drop and the flow rate through an ATU in absence of dust; ii. Establishing or obtaining for the selected filter a second relation between the pressure drop and various amounts of calibrated dust;iii. Establishing a theoretical relation between the time of use of the selected filter and the pressure drop in the ATU, based onA particle concentration obtained from an outdoor air quality indicator, the flow rate settings of the ATU and the second relation between the pressure drop and a amount of calibrated dust established in step (ii); iv. Determining an initial pressure drop AP0for the selected filter from the first relation established in step (i) and the flow rate settings of the ATU, or from a measurement of the pressure drop in absence of dust at the flow rate settings of the ATU; v. Determining a recommended final pressure drop Af by selecting the minimum value between 3 x AP0and AP0+ 100, and optionally choosing a higher or lower value for the final pressure drop; vi. Determining a recommended time of use of the selected filter based on the theoretical relation established in step (iii) and on the recommended final pressure drop Af or the chosen value of final pressure drop A / vii. During utilization of the selected filter in the ATU, measuring an effective pressure drop (Pa) at an effective time of use (h) of the selected filter; viii. Determining a theoretical time of use of the selected filter at the measured effective pressured drop from the theoretical relation of step (iii); ix. Determining a filtration ratio between the theoretical time of use of the selected filter and the effective time of use of the selected filter; x. Determining an adjusted recommended time of use of the filter by dividing the recommended time of use of the filter by the filtration ratio.

[0063] The first relation and / or the second relation may be already provided by the filter manufacturer.

[0064] The expression “effective time of use” refers to the time of use of the filter since the placement of the filter in the ATU and during the operation time of the ATU, thereby not taking into account the time during which the filtration is eventually interrupted. The effective time of use of a filter may be advantageously registered or planned during the operation of the ATU.

[0065] Preferably, in the step of establishing the first relation between the pressure drop and the flow rate through an ATU in absence of dust, at least 3 data points of the pressure drop (Y values) are measured in function of the flow rate (Xvalues), in presence of a clean filter and in absence of dust, wherein the flow rate is varied from 0 m3 / hour to a plurality of values of flow rates representative of the operation of the ATU. The obtained data are preferably fitted to a polynomial equation of the fourth degree.

[0066] Preferably, in the step of establishing the second relation between the pressure drop and a amount of calibrated dust, various amounts of calibrated dust are sent to the selected filter installed in an ATU or in a pilot ATU and the pressure drop is measured at a fixed flow rate, preferably at a flow rate representative of the operation of the ATU. At least 3 data points of the pressure drop (Y values) are measured in function of the dust amounts (X values) and the obtained data are preferably fitted to a polynomial equation of the fourth degree.

[0067] Preferably, the theoretical relation between the time of use of the selected filter and the pressure drop in the ATU is obtained by : calculating a theoretical time of use Ttof the selected filter according to the equation Tt=M‘xl° for each of the amounts of dust used for the establishment of the second relation,- wherein the air volume flow rate qv(in m3 / h) of the ATU is the air volume flow rate selected for operating the ATU and- wherein the PMjnis the particle concentration (in pg / m3) at the inlet of the ATU which can be measured or given by an air quality indicator obtained by querying air quality database based on the location indicator, preferably wherein the location indicator comprises data representative of a geographical location and / or data representative of an environment type associated with the location of the air treatment unit; and reporting each of the calculated theoretical time of use Tt(Y values) in function of their respective pressures drop (X values) and fitting those data in an equation of the fourth degree.

[0068] Preferably, the particle concentration PMjnat the inlet of the ATU is given by the concentration of particles PM2.5 given by the database representative of ageographical location and / or data representative of an environment type associated with the location of the air treatment unit.

[0069] In one embodiment, the user may set up a number of operating hours by day of the ATU, and / or a number of operating days / week of the ATU and ora number of operating weeks by year of the ATU. Advantageously, a monthly average number of effective operating hours may be calculated from the user settings by taking into account the number of operating hours by day, the number of operating days by week and the number of week by year. For example, the user can set up a program of operation of the ATU such that the ATU operates during a number of 14 hours / day, 5 days / week and 52 weeks / years, thereby totalizing an operating time of 3640 hours / year and a monthly average number of operating hours of 303.33 hours.

[0070] In one embodiment, in the theoretical relation between the time of use of the selected filter(in hours) and the pressure drop, the time of use expressed in hours may be normalized by the monthly average number of operating hours of the ATU to obtain normalized effective times values Tn. Therefore in this embodiment, the effective time of use is expressed in months (normalized effective time values Tn) and takes into account that the ATU may not operate full time during the year. The normalized effective time values Tn(Y values) are reported in function of their respective pressures drop (X values) and those data are preferably fitted in an equation of the fourth degree. Then, steps (vii) to (x) are performed wherein the theoretical time, the effective time of use and the adjusted recommended time of use of the filter are expressed in months.

[0071] In another embodiment, the user may set up a program of operation of the ATU according to a calendar. The user may chose to program the ATU to operate between two selected moments of the day (e.g. between 6:00 am and 8:00 pm), during selected days of the year (e.g. all the days of the year or all the day of the year excluding every Saturday, Sunday and public holidays and optionally some other periods of inactivity). The effective time of use (in hours) of the filter in the ATU can also be registered during the operation of the ATU and reported in the calendar. The theoretical time of use and the recommended time of use are expressed in hours and the filtration ratio is determined according to step (ix), then the adjusted time of use (in hours) is determined according to step (x) and the hours of the adjusted time of use are reported in the programmed operation hours in the calendar so that theadjusted time of use of the filter in the ATU takes into account the operating time and interruptions of operations of the ATU.

[0072] In some embodiments, the measurement of the effective pressure drop according to step (vii) can be performed once in a day at a regular time, or several times in the same day.

[0073] In some embodiments, the measurement of the effective pressure drop according to step (vii) can be performed several times in the same day and an average pressure drop of the day is determined based on the different values registered. The average pressure drop of the day may be associated to an effective time of use corresponding to the time of use at the beginning of the day, at the middle of the day or at the end of the day. Then, the theoretical time of use is determined based on the average pressure drop of the day and the filtration ratio is determined by dividing the theoretical time of use by the effective time of use associated to the average pressure drop of the day.

[0074] In some other embodiments, the method of the predictive maintenance may further take into account updates of particle matter in the outdoor air, for example by updating databases of outdoor air quality related to the location of the ATU, or by measuring particle matter concentration at the inlet of the ATU.

[0075] When a plurality of filters is used, the particulate matter concentration arriving to a filter can be estimated based on the PM concentration at the inlet of the ATU and based on the efficiency of filtration of the filters positioned upstream if any. Alternatively PM concentration measurement apparatus can be provided before each filter.

[0076] Additionally, the method of predictive maintenance may comprise a step of providing a energetic cost efficiency indicator of utilization of the filter in the ATU based on the filtration ratio. For example, if the filtration ratio is superior to 1 , for example superior or equal to 1.1 , preferably 1.2, even more preferably to 1.5, it means that dust charged on the filter is accumulated too fast, that associated energy costs are high and that corrective measure to the ATU has to be applied, such as for example correction of the air flow rate.Example

[0077] An optimal air filter was determined for an air treatment unit with operating data as indicated in Table 1. Based on the location indicator, the WHO Ambient Air Quality Database (6thedition, 15 May 2023) is queried to obtain that the PM concentration corresponding with Zagreb is 20 pg / m3(PMjn). Based on the desired supply air quality SLIP3, it is determined that the air filters must meet ISO ePM2.5 of at least 50% (the minimum filtration requirements ISO ePM2,s 50% refers to a final filter stage, according to Eurovent Industry Recommendation 4 / 23 - 2022 Selection of EN ISO 16890 rated air filter classes for general ventilation applications). For purposes of illustrating methods of the present disclosure, two air filters, Filterl and Filter2 of Deltrian International (Fleurus, Belgium) were selected, with operating attributes specified in Table 2. Typically, the initial pressure drop of a filter is given by the manufacturer as a specification of the filter for certain air flow rates passing through a clean filter in absence of dust. In the present case, the initial pressure drop is established for an air flow rate of 3400 m3 / h. For the selected filters Filter 1 and Filter 2, the pressure drop was measured for different masses of dust mi-m2 introduced in a test unit and then, different masses ma-rne to arrive to defined pressures drops in the test unit were weighted. The data thereby obtained were fitted into a polynomial equation of the fourth degree wherein the pressure drop are the Y values and the masses the X values. An end-of-service-life indicator is specified as a condition-based rule with final pressure drop of 250 Pa for both Filterl and 250 Pa Filter2, corresponding to the value given in Table 1 of Eurovent Industry Recommendation 4 / 24 - 2023 for ISO ePM2.s filter classes.

[0078] Based on Eq. 5, the service life TF of the filters can be determined, as indicated in Table 3 and average pressure drop calculated based on Eq. 2 and the operating attributes of Filter 1 and Filter 2. A service life of the ATU is provided as being TA = three years, with a 24 hours per day, 5 days per week and 52 weeks per year of operating time of the ATU. Hence the total electrical energy consumption over TA can be calculated based on Eq. 6. As seen from Table 3, Filter 2, although having a larger electrical energy consumption W per filter for a full service life TF (2282 kWh versus 1513 kWh for Filter 1), will perform better over the entire lifetime TA of the ATU (WT= 5185 kWh versus 5384 kWh for Filter 1)..Table 1 : Operating data of the ATUTable 2: Attributes of the selected air filtersTable 3: Calculation results

[0079] The outcome can further be optimized by adapting TF based on a predetermined logic and calculating electrical energy consumption again. In Table 4, optimization results are obtained when the value of r is reduced to zero by incrementing the value of R (No. of filters utilized over TA) to the next integer. This means that all the filters utilized over the time span TA are utilized over their full service life. By so doing, the total energy consumption can be reduced compared to Table 3. Yet further optimization logic can be applied. By way of example, the EOL indicator can be adapted to increase or reduce the service life of the filter and the total energy consumption calculated and compared with previous results. These optimization strategies can be applied in combination with cost attributes to weigh the effect of increased number of filter replacements compared to lower operating expenses (energy consumption), or even environmental cost attributes when environmental cost of e.g. filter production and disposal are taken into account to calculate and optimize a total carbon footprint.Table 4: Optimization resultsExample of Predictive maintenance.

[0080] An ATU is located in Marseille, France and is set up with an air flow rate of 3400m3 / h and programmed to operate 14h per day, 5 days per week and 52 weeks per year, totalizing an average monthly operating time of 303 hours. By querying relevant database related to the particle concentration in the city at a certain moment, a concentration of PM2,s of 11.1 pg / m3is obtained, corresponding to a class of outdoor air quality ODA3. It is desired by the user to supply an indoor air quality SLIP3. Based on these information, a computer implemented program provides a recommendation of minimum filtration efficiency ePM2,s of at least 80% and provides to the user a selection of a plurality of filters meeting this requirements of filtration efficiency. In some embodiments, a further decision indicator for selecting a filter based on energy consumption can be provided to the user. Additionally, further decision indicators as described in relation of the first aspect of the invention can be provided to the user.

[0081] An appropriated filter 3 has been selected and its specifications are retrieved from a database accessible from the computer implemented program, the database comprising :A first relation between the pressure drop (in Pa) and the flow rate (in M3 / h) through an test ATU in absence of dust; andA second relation between the pressure drop (in Pa) and various amounts of calibrated dust (in g).

[0082] In the selected filter 3, data of measured values of pressure drop in function of the flow rate and in function of various amounts of calibrated dust are shown in table 5. The first and second relations are fitted into polynomial equations of the fourth order.Table 5: specifications of the selected filter

[0083] Then in function of the concentration of PIVk.s (11.1 pg / m3), the flow rate of the ATU (3400 m3 / h), each of the masses used for establishing the second relation are transformed in time of use according to Eq. 5. The computer implemented program establishes a theoretical relation between the time of use (in hours) of the selected filter and the pressure drop in the ATU and fit these data into a polynomial equation of the fourth degree.

[0084] Based on the flow rate of the ATU (3400m3 / h) and the polynomial equation of first relation, the initial pressure drop of the ATU is calculated as 108 Pa. The recommended final pressure drop is then calculated as the minimum value between 3 times the initial pressure drop and the initial pressure drop plus 100, so that the recommended final pressure drop is 208 Pa. The user may select another value of the recommended final pressure, as long as it does not overcomes amaximum of the polynomial function of the pressure drop in function of the mass introduced in the ATU. In the present example, the recommended final pressure drop has been set up by the user at 218 Pa.

[0085] Based on the recommended final pressure drop and the polynomial equation of the theoretical relation between the time of use (in hours) of the selected filter and the pressure drop in the ATU, a recommended time of use of 11548 hours has been calculated. The recommended time of use can be normalized by the average monthly operating time of 303 hours to obtain 38,1 month.

[0086] During operation of the ATU, the pressure drop is regularly monitored. At an effective time of use, a value of an effective pressure drop is registered and this value is used in the equation of the theoretical relation to determine a theoretical time of use corresponding to the registered effective pressure drop. In the present example, the computer implemented program further normalize the time of use (in hours) by the average monthly operating time to obtain data of pressure drop in function of a time of use expressed in months, and those data are fit in a further polynomial equation of the fourth degree.

[0087] In one example of utilization of a filter in an ATU, a new filter 3 was provided on 10 January 2022 and an effective pressure drop of 180 Pa was measured on 2ndaugust 2024. The effective time of use is of 31.2 months and the calculated theoretical time corresponding to a pressure drop of 180 Pa is of 27 months. A filtration ratio of the theoretical time of use on the effective time of use is of 0.87. The recommended date of change of the filter, obtained from the recommended time of use of the filter is on 25 February 2025. The adjusted recommended time of use of the filter is calculated by dividing the recommended time of use of the filter by the filtration ratio, thereby adjusting the recommended date of the change of the filter to 21 August 2025.

[0088] In another example, the same ATU in Marseille, France with the same air flow rate of 3400 m3 / h is considered with the same outdoor concentration of PM2,s of 11.1 pg / m3(ODA3) and the same desired indoor air quality SUP3. The same filter 3 is used in the ATU, so that: the specifications of the filter,the first relation between the pressure drop and the flow rate through a test ATU in absence of dust, the second relation between the pressure drop and various amounts of calibrated dust; the theoretical relation between the pressure drop and the time of use expressed in hours; the recommended final pressure drop (modified by the user to 218 Pa instead of 208 Pa); and the recommended time of use of the filter of 11548 hours, are the same as the previous example.

[0089] In the present example, the computer implemented program is configured to take into account the operating time of the ATU and the inactive time of the ATU. For example, the computer implemented program may allow to the user to set up that the ATU does not operates during weekends, public holidays (e.g. Easter Monday, 1stof May, Ascension Day, Pentecost Monday, Assumption, All Saints’ Day and Armistice) and a certain period of the year such as the week of Christmas and the week of New Year. The computer implemented program further allows the user to set up the time of operation per working day of the ATU, for example 14 hours, so that the program can calculate the cumulated operating hours of the ATU by taking only into account the operating days of the ATU. For example, for a filter 3 installed on 10 January 2022 with a recommended time of use of 11548 hours, the recommended date of change of the filter calculated by the program taking into account the operating days is 23 April 2025.

[0090] When a pressure drop of 180 Pa is measured at the end of the day of 2ndAugust 2024, the effective time of operation of the filter is of 9114 hours. The theoretical time of use of the filter based on the theoretical relation of the pressure drop in function of the time of use expressed in hours is of 8201 hours. The filtration ratio is of 0.899 and the adjusted recommended time of use is calculated from the recommended time of use of 11548 hours divided by the filtration ratio of 0.899, so that the adjusted recommended time of use is of 12833 hours which corresponds to an adapted date of change of the filter of 4 September 2025.

Claims

CLAIMS1. Computer-implemented method of determining a decision indicator associated with an air filter of an air treatment unit (10), the decision indicator allowing to select an optimal air filter (122) for the air treatment unit out of a first plurality of air filters, wherein each of the first plurality of air filters is associated with an operating attribute (216) representative of data indicative of a flow resistance of the respective air filter associated with a clogging state of the air filter, wherein the air treatment unit feeds dust laden air to the air filter to obtain filtered air having a lower particulate matter concentration relative to the dust laden air, the method comprising: providing (201) operating data representative of operating conditions of the air treatment unit, the operating data comprising a volumetric air flow rate indicator (211) of the respective air filter; and an energy efficiency indicator (212) of the air treatment unit, providing (202) a first air quality indicator (215) of the dust laden air, determining (204) an end-of-service-life indicator (EOL) for each of the first plurality of air filters, and determining (205, 206) for each of the first plurality of air filters an electrical energy consumption (W, WT) associated with the air treatment unit based on the volumetric air flow rate indicator (211), the energy efficiency indicator (212), the end-of-service-life indicator (EOL), the operating attribute (216) and the first air quality indicator (215), wherein the decision indicator is based on the electrical energy consumption, characterised in that the first air quality indicator (215) is representative of a particulate matter concentration of the dust laden air at a location of the air treatment unit.

2. Method of claim 1 , wherein determining the electrical energy consumption (W, WT) comprises determining for each of the first plurality of air filters a service life based on the end-of-service-life indicator (EOL) and possibly the first air quality indicator (215).

3. Method of the preceding claim, comprising determining (206) for each of the first plurality of air filters a total electrical energy consumption(WT) over a predetermined operational time period of the air treatment unit, wherein the predetermined operational time period is greater than or equal to the service life, preferably wherein the decision indicator is based on the total electrical energy consumption.

4. Method of any one of the preceding claims, further comprising providing a location indicator (214) representative of a location of the air treatment unit (10) and wherein the first air quality indicator (215) is obtained by querying an air quality database (44) based on the location indicator (214) to retrieve the first air quality indicator (215), preferably wherein the location indicator (214) comprises data representative of a geographical location and / or data representative of an environment type associated with the location of the air treatment unit.

5. Method of any one of the preceding claims, wherein the first air quality indicator (215) is obtained based on measuring an air quality of the dust laden air fed to the air treatment unit.

6. Method of any one of the preceding claims, further comprising determining, for each of the first plurality of air filters, an operating cost of the air treatment unit based on an energy cost attribute (219) and on the electrical energy consumption or optionally on the total electrical energy consumption, preferably wherein the decision indicator is based on the operating cost.

7. Method of claim 6 in conjunction with claim 4, comprising obtaining the energy cost attribute (219) by querying a database based on the location indicator (214).

8. Method of claim 6 or 7, further comprising determining for each of the first plurality of air filters a total expenditure cost (TCO) based on the operating cost and on one or more filter cost attributes, preferably wherein the filter cost attributes are economical or environmental cost attributes comprising one or a combination of: a cost relating to acquisition and / or production of the respective air filter, a cost relating to replacement of the respective air filter and a cost relating to disposal of the respective air filter, preferably wherein the decision indicator is based on the total expenditure cost.

9. Method of any one of the preceding claims, further comprising providing a second air quality indicator (213) of the filtered air and selecting the first plurality of air filters from a second plurality of air filters based on the second air quality indicator.

10. Method of any one of the preceding claims, further comprising: selecting the air filter of the first plurality of air filters corresponding with an extremum of the decision indicator as the optimal air filter.

11. Method of any one of the preceding claims, wherein the decision indicator is representative of an economical cost, an environmental cost or a combination of both.

12. Method of any one of the preceding claims, wherein the dust laden air is outdoor air and wherein the filtered air is supplied to an indoor environment, particularly of a building.

13. Method of operating an air treatment unit (10) comprising an air filter (122), wherein the air treatment unit feeds dust laden air to the air filter to obtain filtered air having a lower particulate matter concentration relative to the dust laden air, the method comprising installing an optimal air filter selected based on a decision indicator determined according to the method of any one of the preceding claims in the air treatment unit and operating the air treatment unit with the optimal air filter, preferably comprising determining a service life associated with the optimal air filter based on the end-of-service-life indicator (EOL) and possibly the first air quality indicator (215), wherein the optimal air filter is replaced at an end of the service life.

14. Data processing system, comprising means for carrying out the method of any one of the preceding claims.

15. Computer program product, comprising instructions which, when the computer program product is executed by a computer, cause the computer to carry out the method of any one of the claims 1 to 12.

16. Method for predictive maintenance of a filter of an air treatment unit (ATU), the method comprising the steps of : establishing or obtaining for a selected filter, a first relation between the pressure drop and the flow rate through an ATU in absence of dust; establishing or obtaining for the selected filter a second relation between the pressure drop and various amounts of calibrated dust; establishing a theoretical relation between the time of use of the selected filter and the pressure drop in the ATU, based on■ A particle concentration obtained from an outdoor air quality indicator,■ the flow rate settings of the ATU and■ the said second relation; determining an initial pressure drop AP0for the selected filter from the first relation and the flow rate settings of the ATU, or from a measurement of the pressure drop in absence of dust at the set flow rate of the ATU; determining a recommended final pressure drop Af by selecting the minimum value between 3 x AP0and AP0+ 100, and optionally choosing a higher or lower value for the recommended final pressure drop; determining a recommended time of use of the selected filter based on the theoretical relation and on the recommended final pressure drop APy or the chosen value of recommended final pressure drop A / during utilization of the selected filter in the ATU, measuring an effective pressure drop (Pa) at an effective time of use (h) of the selected filter; determining a theoretical time of use of the selected filter at the measured effective pressured drop from the theoretical relation of step (iii); determining a filtration ratio between the theoretical time of use of the selected filter and the effective time of use of the selected filter; determining an adjusted recommended time of use of the filter by dividing the recommended time of use of the filter by the filtration ratio.

17. Method according to claim 16, further comprising a step of providing a energetic cost efficiency indicator of utilization of the filter in the ATU based on the filtration ratio.

18. Data processing system, comprising means for carrying out the method of claim 16 or 17.

19. Computer program product, comprising instructions which, when the computer program product is executed by a computer, cause the computer to carry out the method of claim 16 or 17.

Images