Emissions monitoring system

Abstract

There is provided a controller for a multi-emitter monitoring system, the controller being arranged in use to identify an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of the plurality of emitters; and arranged in use to instruct monitoring of at least one of the plurality of emitters based on the identified importance. Corresponding methods, systems and computer programmes are also provided.

Description

[0001] EMISSIONS MONITORING SYSTEM

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to the monitoring of emitters. This is intended to be for emissions, such as greenhouse gas emissions monitoring purposes, typically of multiple maritime and land-based combustion engines.

[0004] BACKGROUND

[0005] Reporting on industrial emissions has historically been based on monitoring surveys or campaigns. This involves attaching mobile equipment to a system and monitoring the emissions of the system.

[0006] The campaigns and surveys typically run over a period of several days but are, typically, limited to less than a week. The data recorded is then processed to identify the emissions of the system.

[0007] This is typically used to represent the emissions of the system at all times. However, due to the limited length of the campaign or survey, this is not considered to be accurate or representative of how such systems operate and what they emit under normal operating conditions. This is because the monitoring will cover all the conditions experienced by the system, meaning variations in emissions due to conditions not occurring during the monitoring will not be captured.

[0008] The equipment used for campaigns and surveys is typically a continuous emissions monitoring system (CEMS). With sufficient data handling capacity, it is possible to conduct monitoring at all times through the use of CEMS.

[0009] A CEMS is connected to an output, such as an exhaust flue, of an emitter allowing exhaust to be sampled by the CEMS. CEMS typically include an analyser to which samples are passed for processing to identify the desired readings.

[0010] Each analyser is only able to process one sample at a time. However, a large number of systems have multiple emitters. To provide full-time monitoring of multiple emitters at the same time, multiple CEMS, or at least multiple analysers, are needed. This is often impractical due to limitations on available space and can be prohibitively expensive.

[0011] We have developed a means of monitoring multiple emitters while limiting the space required and the number of analysers required. This has been achieved by cycling from one emitter to the next. However, as the number of emitters increases, this lengthens the amount of time between periods during which a respective emitter is monitored.

[0012] As an example, the instruments used to analyse the exhaust can have a 2% error rate. If sampling of each emitter in the group is conducted once every five minutes instead of continuously, we have found the error rate is increased to around 4% to 5%. Since the analysis is non-continuous, averaging of the results is then needed, for which the error rate must be considered by applying mean squared errors. We have found this gives an error rate of around 20% to 25% depending on the number of emitters.

[0013] This significant increase in error rate of about 2% for continuous monitoring to about 20% for non-continuous, but regular monitoring in a cycle is again prohibitive. There is therefore a need to be able to allow monitoring of multiple emitters while maintaining accuracy and limiting monitoring equipment requirements and units required.

[0014] SUMMARY OF INVENTION

[0015] According to a first aspect, there is provided a controller for (i.e. suitable for) a multi-emitter monitoring system, the controller being arranged in use to identify an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of (each of) the plurality of emitters; and arranged in use to instruct monitoring of at least one of the plurality of emitters based on the identified importance.

[0016] When using a controller according to a first aspect, instead of cycling through emitters in a regimented sequence, this allows the most relevant emitter to be scheduled for monitoring before other emitters. Further, this allows the importance of monitoring each emitter to be factored into the decision on which emitter to monitor. Fundamentally, this provides improved monitoring accuracy through the adaptive selection of emitters for monitoring.

[0017] By the term “instruct” we intend to mean that monitoring of the relevant emitter is to be initiated immediately, next after monitoring of any emitter currently being emitted is complete. This is intended to include immediately after monitoring of any emitter currently being monitored is completed, or after a switching time and / or clearance time has elapsed after monitoring of any currently monitored emitter has passed.

[0018] In preferable examples, the controller is arranged to instruct monitoring by an analyser connected to each of the plurality of emitters, and the controller is arranged in use to instruct the analyser to monitor at least one of the plurality of emitters. The analyser may be a single analyser, where the analyser is connected to multiple emitters. In such examples, a sample for analysis may be transported from the emitter to the analyser, as is discussed in detail below.

[0019] Similarly, instructing monitoring of at least one of the plurality of emitters may comprise: instructing an analyser to monitor a first emitter of the plurality of emitters; and instructing the analyser to monitor a second emitter of the plurality of emitters.

[0020] It is intended that the term “emitter” is any object, system, device or apparatus that provides an output. Typically, however, this is intended to be an object, system, device or apparatus that issues waste or exhaust, that emits byproducts as part of a process of providing a primary output or function, or that issues a substance. This is intended to include engines or motors, such as combustion engines, but may include furnaces, milling, heat treating, refining or other industrial production systems. Regardless of the form the emitter takes, the relevant output may be greenhouse gas (GHG) contributing output. The at least one attribute of the plurality of emitters may include at least one activity parameter of each of the plurality of emitters and / or at least one parameter of each of the plurality of emitters. This links the importance to activity or the emitters and / or other factors, such as physical or operational parameters of the emitters. This means the monitoring is linked to activity and / or physical / operational emitter parameters rather than being based on non-emitter-based matters.

[0021] When identifying the importance of monitoring each emitter of the plurality of emitters, the controller may be arranged in use to receive at least one activity parameter of each of a plurality of emitters of a multi-emitter system. This links the monitoring to the activity of the emitters allowing monitoring to be adjusted in view of which emitters are active and which are inactive, for example.

[0022] In another form, in implementing identification of the importance, the controller may comprise: an input arranged in use to receive (the) at least one activity parameter.

[0023] The at least one activity parameter can be any of several parameters. Typically, the at least one activity parameters include (i.e. each at least one activity parameter includes) an indicator, the indicator identifying if the emitter is active or inactive. This allows tailoring of the monitoring to favour the active emitters over inactive emitters, such as to only monitor or schedule monitoring of active emitters, or to allow higher rankings of active emitters than inactive emitters. This makes monitoring more efficient and reduces the amount of time between occasions when active emitters are monitored.

[0024] In a similar manner, the at least one attribute may be an indicator, the indicator identifying if the emitter is active or inactive.

[0025] Typically, the indicator may be a representation of one or a combination of a Boolean value, an emitter rate threshold and an emitter load measure. These factors allow for a simple means of identifying if each emitter is active or inactive. For example, the Boolean value is able to be an identifier of a respective emitter being active, i.e. on, or inactive, i.e. off or idle. This may be provided from the emitter or from any entity operating the emitter.

[0026] The emitter rate threshold may be a minimum use rate, such as a minimum flow rate of emitter output, such as exhaust output flow rate. Example thresholds include a exhaust flow rate of at least 4 kilograms per second (kg / s) or at least 1 kg / s. The threshold may be a threshold to which fuzzy logic is applied since this allows edge cases where the emitter rate is fluctuating at about the threshold level.

[0027] The emitter load measure may be a measure of the load on the emitter or load the emitter is working at. Example emitter load measures include the emitter being in the range of 10% to 25% engine load and / or 25% to 75% engine load.

[0028] Additionally, or alternatively, the indicator may be a representation of one of, or a combination of, rate of change of any non-binary property that is continuous in the temporal domain. For example, a rate of change of (exhaust) flow rate with respect to time (such as dF / dt where F represents flow rate and t represents time).

[0029] When identifying the importance of monitoring each emitter of the plurality of emitters, the controller may be arranged in use to identify a priority score of each of the plurality of emitters based on at least one parameter of each of the plurality of emitters. Typically, the at least one parameters are physical or operational properties of the emitters, but can be environmental or input properties for the emitters. This allows the factors affecting the operation of the emitters to be taken into account when arranging monitoring.

[0030] The at least one parameters may be one or more of several parameters. Typically, the at least one parameters includes the at least one activity parameters. This may be limited to the at least one parameters including only the at least one activity parameters. This allows the identified importance or priority score to be based only on whether the emitters are active or inactive or taking into account the activity parameters in the importance or priority scores. For example, an emitter with an activity parameter that indicates the emitter is active may be given a higher importance or priority score than an emitter with an activity parameter that indicates the emitter is inactive. However, in some circumstances, this may be weighed against other factors and may simply contribute to the importance or priority score rather than determine it.

[0031] Typically, the at least one parameters or attribute may (further) include one or more of: emitter power; emitter Maximum Continuous Rating (MCR); emitter load; emitter operation mode; emitter resource consumption rate; emitter specific fuel consumption (SFC); emitter methane CH4(g / kWh); emitter output temperature; emitter output pressure; and emitter output rate. These parameters allow for the importance or priority score to take account of factors that affect how important it is to monitor each emitter.

[0032] The emitter power or MCR may be an absolute measure for each respective emitter. However, the emitter power or MCR may be relative compared to the other emitters of the plurality of emitters. For example, there may be one or more “main” or “primary” emitters and one or more “auxiliary” or “secondary” emitters with each main emitter being likely to have a higher importance or priority score than each auxiliary emitter.

[0033] The emitter load may be consistent with the emitter load measure detailed above.

[0034] The emitter operation mode may be based on the resource used by the emitter or the means by which the emitter functions. For example, the emitter operation mode may be the fuel type currently being used by the emitter, such as diesel, biodiesel, heavy fuel oil, marine diesel oil, gas (for example, liquefied natural gas), ammonia, methanol, hydrogen or some other fuel. Each fuel type causes the emitter to operate in a different way, and thus a different mode.

[0035] The emitter resource consumption rate may be fuel use rate or some other resource use rate, such as oil or coolant.

[0036] The emitter output temperature may be the temperature of exhaust produced by the emitter, operating temperature of the emitter (i.e. temperature of the physical emitter) or temperature of the output of the emitter. Typically, the exhaust may be exhaust gas produced. Further, or alternatively, the temperature may be a temperature of one or more liners, cooling oil or another attribute. These are often not monitored, however, and may have little relevance to exhaust composition. Instead, the temperature may be the exhaust temperature, such as the exhaust gas temperature.

[0037] The emitter output pressure may be the pressure of exhaust produced by the emitter, or pressure of the output of the emitter. The pressure may be gauge pressure, for example.

[0038] The emitter output rate may be the exhaust, byproduct or product output or flow rate of the emitter.

[0039] The controller may be further arranged in use to rank each of the plurality of emitters based on the activity parameters and the priority scores. This may be a form of the identified importance, or may contribute to the identification of the importance.

[0040] In another form, such as when implementing instruction of monitoring, the controller may comprise: a scheduler arranged in use to rank each of the plurality of emitters based on the activity parameters and the priority scores and may be further arranged to instruct monitoring of a highest ranked emitter.

[0041] The scheduler may only instruct monitoring of a highest ranked or highest identified importance emitter. Typically though, the scheduler is arranged to instruct monitoring of each emitter in rank or importance order. This allows all the emitters to be monitored, but in order of how important it is to monitor the emitter. This schedule may be fixed once the ranking or importance is identified, but, typically, the ranking or importance is able to be adjusted from time to time. The ranking or importance may be adjusted after each of the emitters has been monitored after the rankings or importance are / is identified, or the ranking or importance may be adjusted after each emitter is monitored, or, while one emitter is being monitored, the ranking or importance of all the other emitters may be adjusted. Any adjustments to the rankings or identified importance may be based on the activity parameters and the priority scores (in cases where rankings are used).

[0042] The ranking (or identified importance or attribute) may be (further) based on a time since each emitter was monitored. This provides a means of limiting a period since an emitter was last monitored, thus reducing errors in the monitoring. This may be included in the ranking or importance only for active emitters. In such circumstances, inactive emitters may have a lower priority due to not including a contribution of a period since last monitoring or may be removed from the ranking entirely or ranked as unranked due to being inactive.

[0043] The ranking or importance may be a simple scoring of emitters relative to each other or on an absolute scale, or may be some other means of ordering the emitters. Typically, the ranking or instructing of monitoring is a monitoring timetable, the controller being arranged (such as by the scheduler being arranged) to instruct monitoring of the plurality of emitters based on the timetable. This simplifies how monitoring is to be instructed since it allows monitoring to be attributed to a time or time period and identifies when one emitter is to be monitored relative to one or more other emitters. As noted above in relation to adjusting of rankings, the monitoring timetable is able to be updated or adjusted based on the same or corresponding criteria.

[0044] Regardless of the form of the ranking or instruction monitoring (such as whether the ranking is a score, timetable or some other ordering means), typically, the ranking may be further based on a minimum monitoring frequency. This provides improved accuracy by limiting the time between occasions on which each emitter is monitored. This also reduces the likelihood of missing activity of significant relevance.

[0045] A minimum monitoring frequency may be at least once every 1, 2, 3, 4, 5, 10 or 15 minutes. Typically, the minimum monitoring frequency may be at least once every 4 minutes, which corresponds to a frequency of about 4.16 milliHertz (mHz) The minimum monitoring frequency may be the same for all emitters. Typically, however, active emitters will have at least one minimum monitoring frequency and inactive emitters will have at least one minimum monitoring frequency. The minimum monitoring frequency for active emitters, inactive emitter, all emitters, or each (individual) emitter may be calculated based on the attribute, at least one activity parameters and / or the at least one parameters.

[0046] In some examples, the minimum monitoring frequency may be a minimum regulatory monitoring frequency. In such examples, the controller may ensure compliance with regulatory monitoring intervals by scheduling measurements at frequencies not less than the minimum regulatory monitoring frequency.

[0047] The ranking or importance may be further based on a time (i.e. amount of time or period) since each emitter, or each active emitter and / or each inactive emitter was last monitored. While a minimum monitoring frequency may be a factor in the ranking or importance or not, this means the ranking or importance takes into account the length of time since an emitter was last monitored. The effect on the ranking or importance may vary from emitter to emitter based on the respective at least one parameter and / or respective at least one activity parameter or based on whether the emitter is active or inactive. This means there is less reliance on a minimum monitoring frequency and allows for monitoring, for example, when the time since an emitter was last monitored is longer than the time since one or more other emitters were last monitored. This improves accuracy and limits the possibilities of missing activity relevant for monitoring.

[0048] The use of time since last monitoring, minimum monitoring frequency and / or priority scores may provide an ability for dynamic scheduling. This means the, instead of a fixed “round-robin” monitoring sequence, dynamic adjustments to a monitoring schedule are able to be achieved to tailor monitoring to enhance accuracy and provide monitoring when and where monitoring is best applied. In some examples, non-operating emitters are identified, and monitoring time is reallocated from non-operating emitters to operating emitters. This may advantageously improve the sampling frequency for active emission sources whilst maintaining timing constraints for each individual measurement. When a monitoring timetable is used, the timetable may be based on one or more of: sample travel time; switching duration time; and sample clearance time. As an alternative to sample travel time and sample clearance time, the timetable may be based on a guard period and / or a measurement period. The guard period may be the combination of sample travel time and sample clearance time. The sum of the guard period and measurement period would correspond to a switching duration time.

[0049] By the term “sample” we intend to mean a portion of the emitter output that is monitored, such as gas or greenhouse gas.

[0050] A sample travel time may be the time taken for the sample to travel from a location or point where it is identified at the emitter or an output of the emitter and the location or point at which the sample is monitored. For example, the sample travel time may be the time required for a gas sample to flow through a sample conduit from a respective emitter stack to the analyser.

[0051] A switching duration time may be a time taken for switching to occur, such as for a valve to change from one state to another or for a valve to operate. For example, the switching duration time may be a time required for valve operations to switch the analyser between different emitter stacks.

[0052] A sample clearance time may be a time it takes a sample or any remains of the sample to leave, dissipate or be removed from where it is being monitored from the point or time at which switching occurs. For example, the sample clearance time may be a guard period required after switching. Such a guard period may be used to purge the analyser of a previous sample and / or used to allow a signal from the analyser to stabilise before introducing a sample. For example, the sample clearance may comprise one or more of: gas purging phase, water vapor stabilisation phase, gas analyser stabilisation phase, thermal and / or pressure equilibration phase. In some examples, the sample clearance time is at least 140 seconds. By taking factors such as these into account, the controller is able to manage switching between emitters and, for example, between inlet and outlet measurements accounting for physical constraints, such as, of valve operations; and required “guard” periods between measurements; and sample travel time for monitoring.

[0053] The controller, in some circumstances by the scheduler, may be further arranged in use to identify a monitoring period when instructing monitoring, the controller or scheduler being arranged in use to vary the monitoring period based on the at least one activity parameters. This allows the controller to increase monitoring length and frequency for active emitters and reduce monitoring length and frequency for inactive emitters. This optimises monitoring and use of shared monitoring equipment by providing this adaptive monitoring rate.

[0054] In some examples, the controller is further connected to a water vapor sensor. The water vapor sensor may be positioned proximal to one or more of the emitters, for example, closer to the emitter than other components of the analyser. I n some examples, the sample is first passed through the water vapor sensor, then through the analyser. In some examples, the controller is connected to a plurality of water vapor sensors, each water vapor sensor corresponding to an emitter to which it is proximally located.

[0055] In some examples, the sample is dried after it is measured by the water vapor sensor. In other examples, the sample may be dried before it is passed to the gas analyser, regardless of the presence of a water vapor sensor.

[0056] In some such examples, the controller is configured to instruct monitoring of an emitter when the water vapor content measured by the corresponding water vapor sensor is within a predetermined range. In some examples, the controller is configured to instruct monitoring of an emitter when the water vapor content measured by the corresponding water vapor sensor has stabilised within a predetermined range, that is, is within a predetermined range for a predetermined period. This may advantageously improve the accuracy of the monitoring by enabling the analyser to compensate for different levels of water vapor content in the environment of the emitters at different times.

[0057] In some examples, when a timetable is used (as discussed above), a period for water vapor stabilization may be comprised in the sample clearance time.

[0058] According to a second aspect, there is provided a system for monitoring a plurality of emitters, the system comprising: a detector arranged in use to identify at least one activity parameter of each of a plurality of emitters of a multi-emitter system; a controller according to the first aspect and arranged in use to receive the at least one activity parameters at an input; and an analyser connected to each of the plurality of emitters and arranged in use, based on instructions from a scheduler of the controller to monitor a highest ranked emitter, to monitor an emitter.

[0059] In some such examples, the system further comprises one or more water vapor sensors located proximal to each of the plurality of emitters, as discussed in greater detail above.

[0060] The system according to the second aspect may be a CEMS.

[0061] According to a third aspect, there is provided a control method for a multi-emitter monitoring system, the method comprising: identifying an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of the plurality of emitters; and instructing monitoring of at least one of the plurality of emitters based on the identified importance.

[0062] Typically, the control method may comprise: receiving at least one activity parameter of each a plurality of emitters of a multi-emitter system; identifying a priority score of each of the plurality of emitters based on at least one parameter of each of the plurality of emitters; ranking each of the plurality of emitters based on the activity parameters and the priority scores; and instructing monitoring of a highest ranked emitter of the plurality of emitters.

[0063] The control method according to the third aspect may implement any one or more of the functionalities described above in relation to the first aspect. According to a fourth aspect, there is provided a monitoring method for a multiemitter monitoring system, the method comprising: identifying at least one attribute of each of a plurality of emitters of a multi-emitter system; applying a control method according to the third aspect using the identified attribute; and monitoring, based on control method instructions at least one emitter of the plurality of emitters.

[0064] This may be implemented by the monitoring method comprising: identifying at least one activity parameter of each of a plurality of emitters of a multi-emitter system; applying a control method according to the third aspect using the identified at least one activity parameters; and monitoring, based on instructions from a scheduler of the controller to monitor a highest ranked emitter, an emitter of the plurality of emitters.

[0065] According to a fifth aspect, there is provided a system comprising a plurality of engines, each of which has an exhaust flue, each exhaust being connected to a system according to the third aspect.

[0066] According to a sixth aspect, there is provided a computer program comprising instructions which, when executed, cause an apparatus to perform the method according to the second or fourth aspect.

[0067] According to a seventh aspect, there is provided a non-transitory computer-readable medium comprising the computer program according to the sixth aspect.

[0068] BRIEF DESCRIPTION OF DRAWINGS

[0069] Example controllers, processes, systems and apparatus are described in detail below with reference to the accompanying drawings, in which:

[0070] Figure 1 shows a schematic of an example system;

[0071] Figure 2 shows a flow diagram of an example process;

[0072] Figure 3 shows a flow diagram of an second example process

[0073] Figure 4 shows a flow diagram of a comparative example process;

[0074] Figure 5 shows a flow diagram of a third example process; and Figure 6 shows a flow diagram of a fourth example process.

[0075] DETAILED DESCRIPTION

[0076] An example system for monitoring a plurality of emitters is generally illustrated at 1 in Figure 1. This includes a plurality of emitters 10, an analyser 20 and a controller 30. In various examples, the system also includes an output in the form of a user interface presented to a user at a terminal 40, such as for meeting monitoring, reporting and verification (MRV) regulations.

[0077] In the example shown in Figure 1, each emitter 10 is an engine 12 to which an exhaust 14 (also referred to as an exhaust flue or a stack) is connected. In other examples, each emitter may be another apparatus that provides an output. In terms of use of the engine or other form of emitter, we intend the example system to be representative of a maritime or land-based installation, such as a ship, fixed maritime installation or platform, factory, manufacturing plant or production facility.

[0078] To allow monitoring, in some examples, the analyser 20 is connected to each emitter 10. This is achieved in various examples by a sample conduit 22 being connected to the stack 14 of each emitter. This is intended to provide a passage between the respective emitter to the analyser. In use, in several examples, this allows samples of exhaust to be transported from the respective stack to the analyser.

[0079] In the example shown in Figure 1 , the analyser 20 is connected to the controller 30. In various examples, the controller is also connected to a cloud processor 50. This is optional, since, in some examples, processing is able to be conducted locally at the controller, a local data centre, or local data storage.

[0080] In some examples, the controller 30 is further, or alternatively, connected to a ship data source 60. The ship data source 60 is provided as an example. In other examples, an alternative data source may be used, such as for a different vehicle or for a different installation, such as a land-based installation. The data source is intended to hold various data associated, for example, with the installation, the emitters of that installation and the operating conditions being applied to the emitters.

[0081] Typically, the data source 60 is one or more Automated Measuring Systems (AMS). This provides real-time data capture and detection, which provides frequent and relevant data collection functionality. Further, in various examples, this data contextualises data sampled at a lower frequency than an AMS source.

[0082] In several examples, the controller 30 has a memory 32 and a process 34. In such examples, the memory is able to hold a computer program or instructions that, when executed by the processor, allow the program to be run.

[0083] The controller 30 is arranged in use, in some examples, to implement an example control method. An example of such a control method is generally illustrated at 200 in Figure 2.

[0084] Overall, the control method example generally illustrated at 200 in Figure 2 seeks to identify 210 an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of the plurality of emitters. Once the importance (also referrable to as an “importance level”) is identified, monitoring of at least one of the plurality of emitters is instructed 220 based on the identified importance.

[0085] A means of implementing this is shown in the example process 300 of Figure 3. In this example, the controller receives 310 at least one activity parameter of each of a plurality of emitters of a multi-emitter system, such as the example multiemitter system 1 shown in Figure 1. The controller 30 also identifies 320 a priority score for each of the plurality of emitters based on at least one parameter of each of the plurality of emitters.

[0086] In some examples, the controller 30 then ranks 330 each of the priority of emitters based on the activity parameters and the priority scores. The controller then instructs 340 monitoring of a highest ranked emitter of the plurality of emitters. Looking at the example control processes of Figure 2 and Figure 3 and at system 1 in more detail, this is intended, in various examples, to allow continuous emissions monitoring. In combination, this means that, in some examples, the controller 30 and analyser 20, with their various inputs, are a continuous emissions monitoring system, CEMS. While the process is able to be carried out using the components identified in the figures and herein, in various examples, each component may be a single component or may be a distributed component spread across a plurality of devices.

[0087] As noted above, the analyser 20 is connected to each emitter 10, in various examples, by connection of the sample conduit 22 to the stack 14. Exhaust passing through each stack from the respective engine 12 is able to pass to the analyser via the sample conduit for analysis. Typically, this is achieved by the sample being drawn into the sample conduit either through a component of the conduit or by the analyser. In various examples, this is implemented by a pump or some other means of establishing a negative pressure or pressure gradient to draw exhaust into the conduit.

[0088] In some examples, one or more sensors are provided at or within the stack (or elsewhere) to provide some or all of the analysis or data recording local to the stack instead or in addition to providing analysis at the analyser 20. Examples of such sensors are a flow sensor, temperature sensor, pressure sensor, (dual) temperature and pressure sensor, and a humidity sensor. These can also be located at other suitable locations in other examples. The system according to an aspect disclosed herein uses known sensors or sensor types to these sensors.

[0089] In some examples, the analyser 20 is arranged in use to conduct monitoring of a received sample, which is, typically, a sample of exhaust. In several examples, where the analyser is part of a CEMS, this includes identifying the composition of the sample. In various examples, the analyser additionally, or alternatively, identifies water content and / or oxygen content of a sample being analysed. The composition, water content and oxygen content are each able to be identified using known measurement processes and equipment. The analyser 20 also includes an internal flow sensor in several examples. While this is able to provide output data on flow rate in some examples, typically, this is, at least primarily, used to identify when equipment within the analyser should be activated or recording data.

[0090] Since it is important to be able to identify attributes of exhaust in each stack 14 independently of exhaust in each other stack, the analyser 20 is only used to analyse a sample from one emitter 10 at a time in some examples. Given this, in various examples, the controller 30 is arranged to instruct switching of monitoring between the stacks.

[0091] In a number of examples, the switching is achieved by switching sampling or sample extraction on or off for one or more stacks 14. As shown in Figure 1, in some examples, this is achieved by the sample conduits 22 being connected to a switch 24 of the analyser 20. This is provided, in various examples, by a manifold or valve arrangement.

[0092] In use, the switch 24 is operated by the controller 30 via an actuating link 28. This provides automated switching based on control signals provided by the controller. While Figure 1 shows the actuating link as a separate (direct) connection between the controller and the switch, in practice and / or other examples, this connection is provided through an interface between the analyser 20 and the controller that may be a common interface for other communication.

[0093] In various examples, when an exhaust sample arrives at the analyser 20, when permitted by the switch, this passes onto a module 28 of the analyser for analysis and processing. In various examples, the module is either a single component or is a plurality of components. Regardless, the output from the module is then fed (typically a data) directly or indirectly to the controller 30.

[0094] To identify the appropriate monitoring to conduct (and thus how to control and instruct the switch to provide passage of exhaust from any one stack 14 to the analyser 20), the controller 30 takes account of various factors in various examples. In some examples, this is an importance that is able to be identified based on at least one attribute of the emitters 10.

[0095] One factor of the attribute(s) is at least one (e.g. one or more) activity parameter of each emitter 10. By this we intend to mean, per emitter, the controller 30 receives at least one activity parameter. This means that, typically, the controller receives a plurality of activity parameters - typically one activity parameter for each emitter. Depending on the example, these are provided in a continuous data stream, intermittently, at regular intervals, on request or as a combination of these. Typically, activity data or activity parameters from sensors and / or the AMS is provided continuously in some examples. In various examples, this provides relevant information for appropriate assignment of priorities. In such examples, activity data from an analyser orCEMS more generally can be provided on a non-continuous basis.

[0096] Each activity parameter is typically received from the data source 60, and thus is provided by a centralised data or records system in various examples. In some examples, each activity parameter is additionally or alternatively received (directly) from a respective emitter 10.

[0097] In several examples, the at least one activity parameter includes a Boolean indicator of an emitter being active or inactive (in other words: on or off). This is typically only needs to provided by the data source 60, but is able to be provided from elsewhere.

[0098] In various examples, the at least one activity parameter includes an indicator of a minimum (or maximum) exhaust flow rate threshold having been reached, or a flow rate (which in some circumstances is then compared to a threshold value or range to identify the flow rate relative to the threshold). This is used to indicate that the emitter 10 is active or inactive due to an emitter being considered active when the flow rate is at least at or above a threshold value or range.

[0099] To avoid edge cases, where the flow rate is crossing the threshold or part of the threshold intermittently or without having a sustained or increasing flow rate, fuzzy logic is applied to a flow rate based activity parameter in some examples. This information is provided to the controller 30 by the data source 60 in some examples. In other examples, this information is provided (directly) to the controller from the emitter or another source.

[0100] In some examples, the at least one activity parameter includes a load measure. This can include the emitter 10 having an operational load of, for example, 10% to 25% or 25% to 75% of its maximum. As set out in more detail below, in some situations, this can be paired with a monitoring rate or monitoring frequency to provide dynamic sampling that is load-based. The load rate is provided to the controller by the data source 60 or (directly) from the emitter or another source in a number of examples.

[0101] In other examples, the at least one activity parameters includes or is one or more other factors. Regardless of the specific form of the at least one activity parameters, these provide or contribute to the at least parameter for each of the plurality of emitters 10 on which the controller bases a priority score attributed to each of the plurality of emitters in some examples. This means that in some cases, the priority score can be based on real-time detection of emitter activity and exhaust flow, for example.

[0102] In various examples, additionally, or instead, the priority score is able to be based on at least one (other) parameter for each of the plurality of emitters. In some examples, these include one or more of: emitter size (either as a relative size compared to the other emitters, or as an absolute size); whether a respective emitter is a main emitter or auxiliary; a flow rate of an output (such as exhaust) of the respective emitter, a fuel type used or being used by the respective emitter; a fuel or operating mode of the respective emitter (such as gas or diesel); fuel consumption rate; emitter output temperature; emitter output pressure; emitter output composition; emitter output oxygen content; emitter output water content; emitter efficiency; emitter temperature; emitter operating regime; installation location (such as by being based on GPS data or known fixed location); travel speed and / or direction (where relevant); and one or environmental factors and / or one or more climate or weather factors, such as time of day, precipitation quantity and / or type, ambient temperature, ground temperature, air temperature, sun light levels, cloud cover, wind speed and / or direction. In other examples, a number of further factors contribute to the priority score identification additionally or as alternatives to each other and / or those identified above.

[0103] The priority score, in some examples, is able to be based on an occurrence of a transience regime, such as a change in operating regime. For example, when an emitter switches from inactive to active, such as an idle or inactive engine being brought online or made active, the first 10 to 15 minutes may be highly relevant to the emissions output of that emitter, and so can be of higher importance to monitor. As a further example of a transient property being able to at least contribute the priority score, the priority score is able to be based on a rate of change of parameters or properties, such as a rate of change of exhaust flow rate with respect to time. This and other parameters derived from other input parameters can be used to identify transience regimes. One or more of these can be included as the at least one parameters on which the priority score is based, contributing either as a single parameter or with one or more other parameters.

[0104] The priority score is identified taking into account all of the at least one parameters provided in any one example. This equally applies if the importance being identified based on the at least one attribute is identified by alternative means. Regardless, this is achieved by means of an algorithm in some examples.

[0105] In various examples, the weighting of each parameter of the at least one parameters or attribute is equal, but in other examples, the weightings are varied, variable or tailored to account for the contribution each parameter / attribute as needed, wanted or desired to have on the priority score / importance. For example, emitter size is able to have a high weighting in some examples, compared to emitter output pressure; or exhaust flow is able to have a higher weighting in some examples, compared to location. The algorithm is able to be bespoke to the plurality of emitters for the system 1, or is able to be a generic algorithm that is able to be applied to any system with a plurality of emitters, either generally or in a specific class or group of classes. In some examples, as part of the instructing of monitoring based on importance, each of the plurality of emitters 10 is ranked based on the at least one activity parameters and the priority scores. Of course, when the at least one activity parameters are either the sole contributor to the priority scores or are part of the at least one parameters contributing to the priority score, the at least one activity parameters are accounted for in this manner.

[0106] In various examples, the ranking or instructing based on importance is a timetable according to which some or all of the plurality emitters 10 are to be monitored by the analyser 20 and CEMS. In other examples, the ranking is a simpler ranking of highest to lowest ranked.

[0107] In several examples, the ranking or importance level takes into account the time since an emitter 10 was last monitored. This is achieved in some examples by this being a factor in the priority score, but in other examples, such as when importance is based on one or more different attributes, is independent of the priority scores.

[0108] Typically, the more time since an emitter was last monitored, the higher the priority score, the higher the importance level, is for monitoring again. However, in some circumstances, such as depending on the at least one activity parameter of an emitter or when the emitter 10 is inactive or has been inactive since it was last monitored, the effect on the priority score or importance level for the respective emitter is lower in some examples.

[0109] In a similar manner, in various examples, a minimum monitoring frequency is applied. This is achieved in some examples by the controller having a minimum monitoring frequency as a pre-determined value or requirement. Alternatively, this is varied in a number of examples. Further, in some circumstances, the minimum monitoring frequency for an emitter 10 is different depending on the at least one activity parameter of the respective emitter. For example, the minimum monitoring frequency is lower for an inactive emitter than for an active emitter, such as a monitoring frequency of zero for inactive emitters. This would limit resource requirements for monitoring. While, in some examples, the monitoring order may be fixed once instructed, in other examples, the monitoring order is dynamic. By this, we intended to mean that the monitoring order (i.e. the order in which the emitters are monitored) is able to be changed.

[0110] This is achieved in several examples by (re)assessing the importance, priority score or ranking. In various examples, this occurs at regular time intervals; continuously (either with a delay or in real-time); after a monitoring period of one or more emitters is complete; or before starting monitoring of an emitter. This (re)assessment takes into account the priority scores, (activity) parameters, attributes and importance, and, in some examples, the time since each emitter was last monitored and / or the minimum monitoring frequency.

[0111] As set out above, in some examples, an activity parameter of a load measure can be paired with a monitoring rate or monitoring frequency to provide dynamic sampling that is load-based. This is achieved in various examples by monitoring at a preset frequency when the emitter load is within a predetermined range. This is able to be made dynamic in several examples by the frequency being adjusted when the emitter load changes. In a number of examples, monitoring frequency is able to be adapted for other emitters to account for changes to monitoring frequency on an emitter where the monitoring frequency is being adapted to account for changes in emitter load.

[0112] Due to an (i.e. only a single) analyser being shared between several emitters 10, at least one of the stacks 14 is physically separated from the analyser 20 in some examples. This means that it takes time for sample to travel the distance between a respective stack and the analyser. In such examples, the travel time is able to be factored into the monitoring schedule.

[0113] This is achieved by factoring in sample travel time from each emitter into the monitoring scheduling and providing a window before monitoring of an emitter starts to allow time for a sample to travel from one end of a sample conduit 22 at the respective emitter to the other end of the sample conduit at the analyser 20 before a monitoring period starts. Alternatively, the monitoring period can include the travel time, but is (able to be) lengthened to accommodate the travel time and time needed to conduct monitoring instead of only extending long enough to conduct the desired monitoring.

[0114] In various examples, to avoid cross-contamination or (unwanted) mixing of samples, such as between different emitters, a “guard” period is also applied between monitoring being conducted on two emitters. This period allows sample from one emitter to pass out (or be passed out) of the analyser 20 (such as by being returned to a stack 14, for example through a further, not shown, conduit) or dissipate before a sample from another emitter is introduced into the analyser.

[0115] In several examples, the mechanical and physical arrangement of the analyser 20 and sample conduits 22 is also considered. For instance, in some examples, the monitoring schedule is generated taking into account the amount of time one or more valves take to operate. Those valves (not shown) control the physical switching from one conduit providing sample to another conduit providing sample in various examples.

[0116] This switching that, in several examples, takes into account one or more of these factors, can be considered to be intelligent switching. This is because it involves switching that takes into account parameters that affect the analysis process and seeks to accommodate them.

[0117] Once any one sample is received, it is analysed. In some examples, the system implements a "sample and hold" strategy, where the last valid measurement for each emitter 10 is retained until the next monitoring occasion (also referred to as “sampling cycle” of as (sample) monitoring frequency as referred to above). Other strategies would also be possible, such as forms of interpolation, including linear interpolation. It would also be possible to “supersample”, such as by increasing frequency. This can be to match the sample frequency to that of other instruments operating at a higher frequency. In some such examples, a linear fit or a fit of another model between consecutive points is able to be used to provide intermediate data between the points. An example of a process that implements the priority assessment, sampling and holding can be seen by comparison between the two example flow diagrams of Figure 4 and Figure 5. Figure 4 provides a comparative example, whereas Figure 5 provides an example which implements or complements an aspect disclosed herein.

[0118] Figure 4 shows a comparative example process 400 of a fixed sequence for assessing a plurality of emitters. In Figure 4, these are each referred to as a “stack”.

[0119] A first step is system configuration 410. This includes initialising 412 the fixed stack sequence and setting 414 the timing parameters. In the example shown in Figure 4, this is stated as being an active sampling time of 50 seconds (s), a guard time (so a switching and clearance time) of 100 s, a sample rate (so a period over which a sample is taken, of which multiple consecutive samples are taken during each active sampling time) is 5 s, and there are N stacks to sample.

[0120] The next step is the stack sampling cycle 420. This includes starting 422 with one of the N stacks, stack i, collecting 424 raw measurements from a gas (exhaust) sample), of which the raw data 426 per sample includes, in this example, methane (CH4, CH4) measured in parts per million by volume (ppmv), carbon dioxide (CO2, CO2) measured in percentage volume (% vol), carbon monoxide (CO) measured in parts per million (ppm), sulphur dioxide (SO2, SO2) measured in ppm, nitrogen dioxide (NO2, NO2) measured in ppm, nitrogen oxide (NO) measured in ppm, water (H2O, H2O) measured in % vol, pressure measured in milliBar (mbar), temperature measured in degrees Celsius (°C), flow rate measures in cubic metres per hour (m3 / h, mA3 / h) or as a normalised flow, and power measured in kiloWatts (kW).

[0121] Once collected, the gas is collected, gas processing 430 occurs, typically by an analyser. This includes initialising 432 gas objects and converting 434 to volume fractions 434. This conversion includes volume fraction steps 435 of converting ppm to a fraction, converting percentage to a fraction, calculating nitrogen (N2, N2) and updating volume fractions. The gas densities are then calculated 436. This includes density steps 437 of converting the pressure into Pascals (Pa), converting the temperature from Celsius into Kelvin (K), calculating density based on density (p) corresponding to the multiple of pressure (P) and mass (M) divided by the multiple of the universal ideal gas constant (R) and temperature (T). The molecular mass (MM) share is then set. The following step in the gas processing is that the water content of the gas is processed 438 (such as by known techniques). However, in other examples, this may simply correspond to a sensor input. Accordingly, this step may not be present.

[0122] Flow calculations 440 are then carried out. This includes a process to normalise flow 442, calculate 444 volume (vol) flow, and compute 446 mass flows.

[0123] The mass flow calculation 446 includes the flow steps 447 of converting the normal (i.e. normalised) flow to actual flow; carrying out a wet to dry basis conversion; calculating mass flow as volume (V) multiplied by density; and applying Global Warming Potential (GWP) factors.

[0124] The volume fraction 435, density 437 and flow steps 447 are identified and calculated in some examples consistent with and / or using the United Nations Framework Convention on Climate Change (UNFCC) Clean Development Mechanism (CDM) T00IO8.

[0125] The process progresses to time series assembly 450. This includes generating a stack time series 452 for the respective stack and storing 454 results. The results storage includes storing mass flow in kg / h, methane CO2e in kg / h, total CO2e in kg / h, SFC as grams per kiloWatt-hour (g / kWh) (alternatively, this may be CH4in g / kWh), and the state parameters for the respective stack. The process then switches to the next stack 456, returning the process to the stack sampling cycle 420 starting at stack i 422, which, compared to the “i" stated above is i + 1.

[0126] Turning to the example shown in Figure 5, rather than a fixed schedule process, this shows a priority / importance based process 500. As with Figure 4, this process is suitable for assessing a plurality of emitters, which are each referred to in Figure 5 as a “stack. A first step in this process, in some examples, is initialisation 510. This involves starting 512 the sampling system and identifying 514 the timing parameters.

[0127] In various examples, the timing parameters 514 include an active sampling time of 50 s, a guard time of 100 s, a sample rate of 5 s, and a number of samples per stack of the active time divided by the sample rate, so 50 / 5 (50 divided by 5), corresponding to 10 samples per stack in each active period.

[0128] Following initialisation, in several examples, the process progresses to the priority sequencing 520. This includes receiving at least one parameter for the stack, such as at least one activity parameter. In some examples, this includes receiving 522 control parameters, which include engine load percentage, (exhaust) flow rate, engine run (for example, an indicator that the engine is running or operating or, alternatively, that the engine is not running or operating; i.e. the state of the physical machinery) and gas mode active.

[0129] In a number of examples, the control parameters are then evaluated 523, and stack (priority / importance) scores are calculated 524. In some examples, this is achieved in line with the details set out above.

[0130] The stack priorities are then sorted 525 in various examples. In some cases this is followed by a sampling sequence being built 526 based on the stack priorities. In some examples, this corresponds to the ordering or timetable detailed above.

[0131] In several examples, an assessment is then carried out as to whether the priority should be changed 527. This has an outcome of updating 528 the sequence when it is identified that the priority should be changed, which results in the process returning to step of receiving 522 the control parameters, or evaluating 523 the control parameters, or calculating 524 the stage scores; or an outcome of maintaining 529 the sequence when it is identified that the priority does not need to be changed.

[0132] In some examples, the process then moves to the sampling stage 530. This includes identifying 532 the next stack in the sequence, carrying out an active sampling period 534, providing a guard period 536 and moving to the next stack 538, which cycles back to the identifying step. During this process, the priority change assessment 526 may be carried out at any time in various examples, such as after each stack active sampling period, or after all the stacks have been cycled through.

[0133] For each sample or the samples from each active period, the raw measurements are collected similar to step 424 of Figure 4. This typically may include the raw data set out in step 426 of Figure 4. The gas processing 430 and / or flow calculations 440 of the example shown in Figure 4 are then carried out in various examples. This conducts processing to allow attributes to be derived and / or identified from the samples for each stack.

[0134] The sampling process 530 is followed in some examples by a time series assembly 540. In various examples, this includes establishing 542 a stack-wise time series. This includes providing 544 a time series for each stack. In the example shown in Figure 5, this includes a time series 544a for a first stack “ts1 ”, a time series 544b for a second stack “ts2” and a time series 544c for the Nth stack (i.e. an individual time series for each stack).

[0135] In some examples, each time series 544 includes a stack identifier and / or a time stamp. In various examples, each time series also includes data from each sample, such as methane, carbon dioxide and flow rate data, and / or any one or more of the other sample data, gas processing details or flow calculations.

[0136] In several examples, a union of the active stacks is generated 546. Next, in various examples, this results in a consolidated series being generated 548 of time-based measurements attributable to each stack, such as by a stack identifier. As described in more detail below, this data is then able to be processed further outside of the general sample gathering cycle.

[0137] In various examples, the controller 30 or another device acts as an edge computer. In such examples, data aggregation occurs at the edge computer, collecting and, temporarily, storing data locally. This data is then transmitted to the cloud or another location for processing. This implements bandwidth and network-aware logic, such as by multiplexing the data, to optimise resource usage and limit transmission bandwidth.

[0138] In various examples, the system leverages a distributed architecture to optimize performance and data management. This includes the use of edge computing that allows all data collection and initial processing occur on an edge device held locally, such as located on a maritime vessel; the edge system provides the controller and / or a control method to run a dynamic stack selection algorithm and manages the CEMS analyser; and it allows no user interface to provided locally, such as on-board the vessel, although, in some examples, a user interface is provided.

[0139] In examples using the distributed architecture, some examples include cloud computing, which allows all advanced data processing to be performed in the cloud; the cloud system hosts a user interface and reporting modules; and provides scalable computing resources for handling, in some cases, fleet-wide data analysis (allowing comparison of emissions and efficiency metrics across multiple entities, such as one or more vessels, one or more assets and / or one or more emitting plants in a fleet or group and / or comparison of the same engine(s) or emitter(s)).

[0140] In such distributed architecture examples, data synchronisation is also provided in some circumstances. This allows implementation of a robust mechanism for syncing data between edge devices and the cloud, optimised, for example, for maritime connectivity challenges.

[0141] The analyser 20 and the controller 30 have a (data) interface in some examples. In various examples, there are (data) interfaces between the controller and the cloud 50, user terminal 40 and data source 60. These interfaces are provided by Programmable Logic Controllers (PLCs) in use. In several examples, the terminal is connected to the cloud instead of to the controller.

[0142] Regardless of how the interfaces are provided, in various examples, the physical connections between the devices is wired, wireless or a combination of the two. Overall, while emissions monitoring of any form are able to be controlled using the example controllers, control methods and systems described herein. Typically, these are used in monitoring of greenhouse gases (GHGs), such as by providing outputs on Carbon Dioxide and CO2 equivalents (CO2e), such as those derived from methane and nitrous oxide (N2O, N2O), for example, according to the Intergovernmental Panel on Climate Change (IPCC) factors.

[0143] As an example of this, the example controllers, control methods and systems described herein are able to be implemented as part of the Daphne Technology SA PureMetrics™ reporting system and / or as part of the Gold Standard approved carbon credit methodology, “Methodology for Reducing Methane Emissions from Combustion Engine Exhaust”, at times allowing the methodology to be complied with.

[0144] A schematic representation of an example control method for a multi-emitter monitoring system is shown in Figure 6. In this example, the multi-emitter system comprises four emitters: Stack A, with a first timeline 602; Stack B, with a second timeline 604; Stack C, with a third timeline 606; and Stack D, with a fourth timeline 608.

[0145] In this example process, at time T = 0 seconds (0s), a switching process 610a begins to switch an analyser (not pictured) to monitor Stack A. This is because, at time T = 0s, in this example, Stack A is the highest priority emitter known to the scheduler.

[0146] In this example, the priority of each of the emitters is based (in part) on the time since each emitter was last monitored. In the example of Figure 6, Stack C is not the highest priority emitter as it is inactive, and, accordingly, any time allocated for monitoring Stack C (in this example) is dynamically re-allocated to measuring the other emitters.

[0147] This switching process 610a takes 140s in this example and, as an example, comprises switching a mechanical valve to enable sample transfer from Stack A to the analyser. The sample is also transferred during this time, and, in some examples, a water vapor sensor proximal to Stack A makes a measurement of the water vapor at Stack A.

[0148] At time T = 140s, a sampling process 612a begins, where data acquisition is performed, with measurements being made on the sample transferred to the analyser. This sampling process takes 100 seconds in this example.

[0149] At time T = 240s, a second switching process 610b begins to switch the analyser to monitor Stack B, which is now the highest priority emitter known to the scheduler, as Stack A has been recently measured and is therefore of a lower priority.

[0150] At time T = 380s, the second switching process 610b is complete and a second sampling process 612b begins.

[0151] At time T = 480s, the second sampling process 612b is complete, and the method may continue with a third switching process (not shown) to monitor the highest priority emitter. In this example, that may be Stack D, which is not inactive and has not recently been monitored in comparison to Stacks A and B.

Claims

CLAIMS1. A controller for a multi-emitter monitoring system, the controller being arranged in use to identify an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of the plurality of emitters; and arranged in use to instruct monitoring of at least one of the plurality of emitters based on the identified importance.

2. The controller according to claim 1 , wherein the controller is arranged to instruct monitoring by an analyser connected to each of the plurality of emitters, and the controller is arranged in use to instruct the analyser to monitor at least one of the plurality of emitters.

3. The controller according to claim 1, wherein instructing monitoring of at least one of the plurality of emitters comprises:instructing an analyser to monitor a first emitter of the plurality of emitters; andinstructing the analyser to monitor a second emitter of the plurality of emitters.

4. The controller according to any one of the preceding claims, wherein the at least one attribute of the plurality of emitters includes at least one activity parameter of each of the plurality of emitters and / or at least one parameter of each of the plurality of emitters.

5. The controller according to any one of the preceding claims, wherein to identify the importance of monitoring each emitter of the plurality of emitters, the controller is arranged in use to:receive at least one activity parameter of each of a plurality of emitters of a multi-emitter system.

6. The controller according to claim 5, wherein the at least one activity parameter includes an indicator, the indicator identifying if the emitter is active or inactive.

7. The controller according to claim 6, wherein the indicator is a representation of one or a combination of a Boolean value, an emitter rate threshold and an emitter load measure.

8. The controller according to any one of claims 2 to 7, wherein to identify the importance of monitoring each emitter of the plurality of emitters, the controller is arranged in use to:identify a priority score of each of the plurality of emitters based on at least one parameter of each of the plurality of emitters.

9. The controller according to any one of claims 2 to 8, wherein the at least one parameters includes one or more of:emitter power;emitter Maximum Continuous Rating (MCR);emitter load;emitter operation mode;emitter resource consumption rate;emitter specific fuel consumption (SFC);emitter output temperature;emitter output pressure; andemitter output rate.

10. The controller according to claim 8 or claim 9, wherein the at least one parameters includes the at least one activity parameters.

11. The controller according to claim 8 or claim 9, wherein the controller is further arranged in use to rank each of the plurality of emitters based on the activity parameters and the priority scores.

12. The controller according to claim 11 , wherein, to instruct monitoring of at least one of the plurality of emitters, the controller is arranged in use to instruct monitoring of a highest ranked emitter.

13. The controller according to claim 12, wherein the controller is arranged in use to instruct monitoring of each emitter in rank order.

14. The controller according to any one of the preceding claims, wherein the at least one attribute includes a time since each emitter was monitored.

15. The controller according to any one of the preceding claims, wherein the controller is arranged in use to instruct monitoring of the at least one emitter based on a monitoring timetable, the timetable being based on the identified importance of the respective at least one emitter.

16. The controller according to any one of the preceding claims, wherein the controller is arranged in use to instruct monitoring of the at least one emitter based on a monitoring timetable, the timetable being based on the identified importance of the respective at least one emitter, and the timetable is based on one or more of:sample travel time;switching duration time; andsample clearance time.

17. The controller according to any one of the preceding claims, wherein the controller is arranged in use to instruct monitoring of the at least one emitter based on a minimum monitoring frequency.

18. The controller according to any one of the preceding claims, wherein the controller is further arranged in use to identify a monitoring period when instructing monitoring.

19. The controller according to claim 18, wherein the controller is arranged in use to vary the monitoring period based on the identified importance.

20. A system for monitoring a plurality of emitters, the system comprising:a detector arranged in use to identify at least one activity parameter of each of a plurality of emitters of a multi-emitter system;a controller according to any one of claims 1 to 19 arranged in use to receive the at least one activity parameters at an input; andan analyser connected to each of the plurality of emitters and arranged in use, based on instructions from a scheduler of the controller to monitor a highest ranked emitter, to monitor an emitter.

21. The system according to claim 20, wherein the system is a continuous emissions monitoring system (CEMS).

22. The system according to claim 20 or claim 21, further comprising a plurality of water vapor sensors, each water vapor sensor corresponding to and arranged proximal to an emitter of the plurality of emitters.

23. A control method for a multi-emitter monitoring system, the method comprising:identifying an importance of monitoring each emitter of a plurality of emitters of a multi-emitter system based on at least one attribute of the plurality of emitters; andinstructing monitoring of at least one of the plurality of emitters based on the identified importance.

24. The control method according to claim 23, the method comprising:receiving at least one activity parameter of each of a plurality of emitters of a multi-emitter system;identifying a priority score of each of the plurality of emitters based on at least one parameter of each of the plurality of emitters;ranking each of the plurality of emitters based on the activity parameters and the priority scores; andinstructing monitoring of at least one emitter of the plurality of emitters.

25. A monitoring method for a multi-emitter monitoring system, the method comprising:identifying at least one attribute of each of a plurality of emitters of a multiemitter system;applying a control method according to claim 24 using the identified attribute; andmonitoring, based on control method instructions at least one emitter of the plurality of emitters.

26. A system comprising a plurality of engines, each of which has an exhaust flue, each exhaust being connected to a system according to claim 20 or claim 21.

27. A computer program comprising instructions which, when executed, cause an apparatus to perform the method according to claim 23.

28. A non-transitory computer-readable medium comprising the computer program according to claim 27.

Images