TECHNICAL FIELD

The present invention relates to the determination of heat loss in a service line of a district energy system, such as a district heating system or a cooling distribution network.

BACKGROUND

In district energy systems a fluid is transported from an energy supply plant, e.g. a district heating plant, via a grid of fluid lines to a plurality of consumers. To this end, the district energy system comprises a network of supply lines and a plurality of service lines. The service lines branch off from the supply lines and connect the respective consumers with the supply line. The grid of fluid lines for distributing the fluid to a plurality of consumers will also be referred to as district energy grid.

A problem for the district energy grid operator is that, contrary to the supply lines, there are currently no good ways of inspecting the conditions of the service lines. For example, while modern energy grids are equipped with electrical detection systems for detecting leakage or other damages of the supply lines, the service lines typically have no such electrical detection system. Moreover, the service lines are typically located on private property, which creates difficulties when performing thermal inspections or other inspection types.

Knowing the condition of the service lines is important as there are large financial resources bound in a district energy grid, and the grid operator needs to plan maintenance in advance and avoid expensive ad hoc repairs of failed pipes, such as leaking service lines. A fault in a service line can for example be a service line where the insulation has deteriorated or, even worse, a service line leaking district energy water. Therefore, it would be desirable if such an inspection can be done remotely. Moreover it would be desirable when faults in a service line can be detected early, in particular before leakage occurs.

EP3531368 discloses a method for determining a flow rate and a temperature of a fluid at a selected position in a district heating or cooling utility distribution network, comprising a plurality of interconnected distribution lines and smart utility meters. The meters are arranged to register the energy delivered at consumer premises situated along the distribution lines, the method comprising the steps of: collecting meter data time series from the smart utility meters using an Advanced Metering Infrastructure, and calculating the temperature and the flow rate at the selected position. The calculations are based on flow and temperature information derived from the collected meter data time series, the topology of the utility distribution network and heat transfer coefficients of the distribution lines. The meter data time series comprises the accumulated volume of fluid delivered to the consumer, and the integrated flow-temperature product calculated by the meter.

However, the above prior art method requires a detailed network model in terms of pipe diameters, lengths, insulation class, ambient temperature, and heat loss coefficients. For a utility provider with large resources to maintain this information in a GIS database, it may be possible to obtain such detailed information. However, district energy grids have often been in the ground for many years why dimensional and heat loss coefficient data may not always be available, or the district heating grid operator is not maintaining the information to the level needed. This is especially a problem at the service line level.

SUMMARY

Thus, it remains desirable to provide a method and system for determining heat loss in a service line of a district energy system that solve one or more of the above problems and/or that have other benefits, or that at least provide an alternative to existing solutions. According to one aspect, disclosed herein are embodiments of a computer-implemented method for determining thermal loss in a district energy grid, the district energy grid comprising one or more supply lines and a plurality of service lines for transporting a fluid to a plurality of consumers, each supply line feeding the fluid into a plurality of service lines and each service line connecting an consumer with one of the one or more supply lines, the method comprising:

- obtaining measured sensor data from one or more sensors associated with one or more selected service lines of the plurality of service lines;

- determining, from the received sensor data, respective observed fluid temperatures and observed fluid flows of the transported fluid received at respective points in time via respective ones of said one or more selected service lines;

- obtaining a set of equations, the set of equations comprising a plurality of equations relating the observed fluid temperatures and the observed fluid flows with thermal loss parameters indicative of thermal losses of respective ones of said selected one or more service lines, and

- computing a solution of the set of equations and computing a result value of at least a first thermal loss parameter indicative of a thermal loss of a first service line of said selected one or more service lines from the computed solution of the set of equations.

Accordingly, in various embodiments of the method and system disclosed herein, the losses of some or all of the service lines of the district energy grid can be calculated directly from sensor data obtainable from the consumers' heat meters or other suitable logging devices. Embodiments of the method and system disclosed herein do not require detailed knowledge of a network model or the modelling of temperature and flow at respective nodes of the main grid. In some embodiments, each of the selected service lines has a logging device associated with it, such that the associated logging device is configured to record sensor data pertaining to the service line with which the logging device is associated. It will be appreciated that the method may determine thermal energy losses for all or for only some service lines of the district energy grid, e.g. for only a single selected group of service lines or for different selected groups of service lines of the energy grid. Embodiments of the method and system described herein provide the district energy grid operator with information about service pipe losses in sub areas of the grid. Accordingly, the district grid operator can use that information to plan maintenance ahead, even when only modest knowledge of the network is available. Various embodiments of the method disclosed herein allow early detection of faults in service lines, e.g. faulty insulation which might otherwise result in pipe corrosion.

Generally, in various embodiments, the thermal loss parameters of the respective service lines are unknown variables of the set of equations. Computing a solution of the set of equations may thus comprise numerically solving the set of equations for the thermal loss parameters.

In some embodiments, the plurality of equations are linear in the service loss parameters, thus allowing for a computationally efficient solution of the equations.

In particular, in some embodiments, the plurality of equations relate instant heat losses in the respective service lines to the fluid flow in the respective service pipes and a temperature difference between the observed fluid temperatures, observed for the respective service lines, and a fluid temperature in the supply line. The fluid temperature of the supply line may be a further unknown of the set of equations.

In some embodiments, the set of equations is an overdetermined set of equations and the solution of the set of equations is an approximate solution of the overdetermined set of equations. In particular, the number of equations in the set of equations may be larger than the number of thermal loss parameters and, in particular, larger than the number of selected service lines. Computing the solution of the set of equations may thus comprise computing an approximate solution, in particular a least-squares solution, to the overdetermined set of equations. To this end, computing the solution of the set of equations may comprise applying a least-square method for determining the approximate solution of the overdetermined set of equations. In some embodiments, computing the solution of the set of equations comprises computing the solution as a solution to a constrained optimization problem. In particular, in some embodiments, computing the solution of the set of equations comprises minimizing or maximizing a cost function over at least one additional temperature parameter indicative of a temperature and/or a temperature loss associated with the supply line. Alternatively or additionally, computing the solution of the set of equations comprises minimizing or maximizing a cost function over at least one ground temperature parameter that is associated with a ground temperature. Accordingly, the computation of the set of equations does not require prior knowledge of the temperature and/or a temperature loss associated with the supply line and or prior knowledge of the ground temperature.

Preferably, the set of equations includes equations relating to different service lines, thus allowing an assessment of the state of multiple service lines and providing a computed estimate that is less sensitive to noise. Accordingly, in some embodiments the one or more selected service lines include a plurality of service lines, and wherein each of the plurality of equations relates a thermal loss parameter of a respective service line with an observed fluid temperature and an observed fluid flow associated with said respective service line. Preferably, the selected plurality of service lines are all connected to a single supply line, in particular to a section of a single supply line that does not include any branch points other than the service lines branching off to individual consumers. Preferably, the plurality of equations includes, for some or all of the selected one or more service lines, multiple equations associated with respective points in time within an observation period of time, thereby providing a more reliable estimate for the thermal losses during said period of time.

In particular, in some embodiments, the set of equations comprises a first equation relating an observed fluid temperature and an observed fluid flow of the first service line at a first point in time with the first thermal loss parameter. Preferably, the plurality of equations further comprises another equation associated with a second point in time and relating an observed fluid temperature and an observed fluid flow of the first service line at said second point in time with said first thermal loss parameter, the second point in time being different from the first point in time. Alternatively or additionally, the one or more selected service lines include a first and a second service line, different from the first service line, and the plurality of equations comprises another equation, associated with the second service line and relating an observed fluid temperature and an observed fluid flow of said second service line with a second thermal loss parameter indicative of a thermal loss of said second service line. In some embodiments, the first equation and the another equation associated with the second service line each include a temperature parameter associated with the supply line, the temperature parameter representing an unknown variable of the first equation and of the another equation associated with the second service line, i.e. the computational model may exploit the fact that the selected service lines are all connected to the same supply line or otherwise such that the temperature in the supply line is substantially uniform across the selected service lines.

The present disclosure relates to different aspects including the method described above and in the following, corresponding apparatus, systems, methods, and/or products, each yielding one or more of the benefits and advantages described in connection with one or more of the other aspects, and each having one or more embodiments corresponding to the embodiments described in connection with one or more of the other aspects and/or disclosed in the appended claims.

In particular, embodiments of the method disclosed herein may be computer- implemented. Accordingly, disclosed herein are embodiments of a data processing system configured to perform the steps of the method described herein. In particular, the data processing system may have stored thereon program code adapted to cause, when executed by the data processing system, the data processing system to perform the steps of the method described herein. The data processing system may be embodied as a single computer or other data processing device, or as a distributed system including multiple computers and/or other data processing devices, e.g. a client-server system, a cloudbased system, etc. The data processing system may include a data storage device for storing the computer program and sensor data. The data processing system may include a communications interface for receiving sensor data from consumer's heat meters or other suitable logging devices and/or sensors associated with the individual service lines. The data processing system may receive the sensor data from the heat meters and/or other logging devices and/or sensors via a suitable wired or wireless communicative connection, e.g. via a suitable communications network.

The data processing system may provide a user-interface for allowing a user to monitor the computed thermal loss parameters and/or other data associated with the conditions of individual service lines. The data processing system may also issue warnings or alerts or other notifications responsive to detected losses or other conditions, e.g. audible or visual alerts, alerts communicated via e-mail, SMS, or other forms of notifications, and/or the like.

Another aspect relates to a district heating or cooling system. The district heating or cooling system comprises:

- a district energy grid, the district energy grid comprising one or more supply lines and a plurality of service lines for transporting a fluid to a plurality of consumers, each supply line feeding fluid into a plurality of service lines and each service line connecting a consumer with one of the one or more supply lines,

- a plurality of sensors configured to output sensor data indicative of measured fluid temperatures and associated fluid flow of the transported fluid received at respective consumers via the respective service lines,

- a data processing system as described above and in the following.

The sensor data indicative of the measured fluid temperatures and the associated fluid flow may represent the fluid temperatures and associated fluid flow directly. Alternatively, the fluid temperatures and the associated fluid flow may be derivable from the sensor data. For example, the sensor data indicative of the fluid flow may be sensor data directly representing the fluid flow or it may be sensor data from which the fluid flow can be derived, e.g. sensor data representing received volumes of fluid.

Yet another aspect disclosed herein relates to embodiments of a computer program configured to cause a data processing system to perform the acts of the method described above and in the following. A computer program may comprise program code means adapted to cause a data processing system to perform the acts of the method disclosed above and in the following when the program code means are executed on the data processing system. The computer program may be stored on a computer-readable storage medium, in particular a non-transient storage medium, or embodied as a data signal. The non-transient storage medium may comprise any suitable circuitry or device for storing data, such as a RAM, a ROM, an EPROM, EEPROM, flash memory, magnetic or optical storage device, such as a CD ROM, a DVD, a hard disk, and/or the like.

BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments will be described in more detail in connection with the appended drawings, where

FIG. 1 schematically shows an embodiment of a district energy system.

FIG. 2 schematically shows another embodiment of a district energy system.

FIG. 3 shows a flow diagram of an example of method for determining thermal loss in a district energy grid.

FIG. 4 shows a flow diagram of a method for computing estimated thermal loss parameters as a solution of an optimization problem.

FIGs. 5A-F illustrate examples of measured fluid temperatures and flows and examples of results computed by an embodiment of the method described herein.

DETAILED DECRIPTION

FIG. 1 schematically shows an embodiment of a district energy system, generally designated by reference numeral 1. Generally, the district energy system comprises a district energy grid 2, which comprises one or more supply lines 6 and a plurality of service lines 5 for transporting a fluid to a plurality of consumers 4. Each supply line 6 feeds the fluid into a plurality of service lines 5 and each service line 5 connects a consumer 4 with one of the one or more supply lines 6.

The supply lines 6 and/or the service lines may be pipes or other suitable fluid conduits, preferably thermally insulated pipes. The fluid may be water or another suitable medium facilitating heat transfer by advection of the medium through the district energy grid. The district energy system comprises an energy supply plant 3. In the example of FIG. 1, the district energy system is a district heating system and the energy supply plant is a district heating plant which feeds heated liquid into the district energy grid 2.

The district energy grid may comprise a network of supply lines, also referred to as main grid, e.g. as illustrated in FIG. 2. Each consumer 4 is connected to the main grid by a service line 5, e.g. by an insulated service pipe.

Typically, the main grid has one or more supply lines 6 and one or more return lines 7. The supply lines form a supply path for feeding the fluid from the energy supply plant 3 to the individual consumers 4 and the return lines 7 define a return path for returning fluid from the consumers to the energy supply plant 3.

The district energy system further comprises a plurality of logging devices 8. The logging devices 8 are configured to provide temperature sensor data and flow sensor data. The temperature data may be indicative of the measured fluid temperature of the fluid arriving at the respective consumers, also referred to as the supply temperature, In particular, the service line has a receiving end and a consumer end, opposite the receiving end. The service line is fluidly connected to the supply line and receives fluid from the supply line at the receiving end. The consumer receives the fluid via the service line at the consumer end of the service line. The supply temperature represented by the sensor data is indicative of the fluid temperature of the fluid arriving at the consumer at the consumer end of the service line. The flow sensor data may be sensor data directly representing the fluid flow or it may be sensor data from which the fluid flow can be derived. For example, the flow sensor data may represent volume data indicative of the accumulated volume of fluid received by the consumer through the service line. The fluid flow may then be derived as a derivative of the volume data with respect to time. Typically, the logging devices are provided in the form of heat meters, that are installed at the respective consumers 4 and that are configured to log the supply temperature and the accumulated volume of fluid received by the respective consumers. It will be appreciated that, in alternative embodiments, the logging devices may include other types of sensors, e.g. they may be embodied as individual and separate sensors for measuring temperature and volume/flow, respectively.

The district energy system further comprises a data processing system 9, e.g. a server computer, a network of multiple computers, e.g. a client-server architecture, a distributed computing architecture, a cloud-based computing environment, and/or the like. The data processing system 9 may be located at the energy supply plant 3 or at a different location. The data processing system 9 receives logged sensor data from the logging devices 8. To this end, the logging devices 8 at the respective consumers are communicatively coupled to the data processing system 9, e.g. directly or indirectly. The logging devices 8 may communicate the logged sensor data to the data processing system via wired and/or wireless connections, e.g. via a suitable communications network, such as a cellular telecommunications network, via the internet or another suitable computer network. For example, many modern heat meters are configured for automatic communication of the logged data to a central data processing system. The logging devices may communicate the sensor data continuously or intermittently, e.g. one measurement at a time or as batches of multiple, time-stamped sets of sensor data. In some embodiments, the data processing system obtains the logged sensor data from a suitable data repository where the logged sensor data has previously been stored, e.g. by the heat meters, by an energy grid maintenance or surveillance system, or the like.

The data processing system 9 receives the logged sensor data and is programmed to process the received sensor data to determine thermal loss in the district energy grid as described herein. Embodiments of a method implemented by the data processing system will be described in more detail below with reference to FIGs. 3 and 4. The data processing system may store the received sensor data and the results of the determination of the energy loss. Alternatively or additionally, the data processing system may implement functionality for presenting the sensor data and/or the results of the energy loss determination to a user of the data processing system, e.g. in the form of graphs, tables or the like. FIG. 2 schematically shows another embodiment of a district energy system 1. The district energy system 1 of FIG. 2 is similar to the district energy system of FIG. 1 in that it comprises an energy supply plant 3, an energy grid 2 including supply lines and service lines for distributing a fluid from the energy supply plant to a plurality of consumers, a plurality of logging devices 8, e.g. heat meters or other types of sensors, installed at the respective consumers, and a data processing system 9, all as described in connection with FIG. 1. Even though not explicitly shown in FIG. 2, it will be appreciated that the energy grid 2 of FIG. 2 also includes return lines defining a return path as described in connection with FIG. 1. Similarly, even though not explicitly shown in FIG. 2, the logging devices 8 are preferably communicatively coupled to the data processing system 9 and configured to transmit logged sensor data to the data processing system. It will generally be appreciated that, alternatively to communicatively coupling the logging devices to the data processing system, sensor data may be read out and entered into the data processing system in a different manner, e.g. by storing the logged sensor data on a portable data carrier, and by reading the stored data by the data processing system from the portable data carrier. However, it will be appreciated that a communication of the logged sensor data via a suitable communications, link allows a more efficient and, optionally, even (quasi) realtime transmission and analysis of the logged sensor data.

FIG. 2 illustrates that the energy grid 2 typically includes more than a single supply line, in particular a network of supply lines 6 that are interconnected at nodes 20. The nodes 20 typically include valves, which allow the energy grid operator to direct the fluid flow along respective paths through the grid, e.g. so as to temporarily isolate certain parts of the grid from each other for maintenance, or the like. In the example of FIG. 2 open valves are designated V and closed valves are designated C. Accordingly, the energy grid 2 includes multiple subareas 2a - 2e where all consumers in a given subarea are connected to a single supply line. In each subarea of the grid, there are multiple logging devices 8.

Embodiments of a process or determining thermal loss in a district energy grid will now be described with reference to FIG. 3 and with continued reference to FIGs. 1 and 2. The process may be performed by the data processing system 9. For example, some embodiments of the method may be implemented by a cloud computing environment or another suitable data processing system. The determined thermal loss may be calculated and represented as a suitable loss figure of merit indicative of the thermal loss. The loss figure of merit will also be referred to as thermal loss parameter. It may e.g. be output in the form of a graphical indicator of that figure of merit, e.g. via an internet browser or another suitable user interface.

In particular, FIG. 3 shows a flow diagram of an example of a computer-implemented process for determining thermal loss in a district energy grid.

In initial step SI, the process groups the service lines and the associated logging devices 8 of the district energy grid 2 into respective groups. To this end, the process may set up a data structure representing the district energy grid and identifying respective selected groups of data logging devices. The grouping may be performed manually, automatically or in a semi-automatic, user-assisted manner. The grouping is typically only performed once or when changes to the structure of the energy grid occur.

Generally, various embodiments of the process use data from a selected group of logging devices 8 within a geographically limited region of the energy grid 2 where the supply line temperature of the fluid fed from the main district energy grid into the service lines can be expected to be roughly the same for all service lines or where differences in the supply line temperature between users can easily be modelled. In the following, such a geographically limited region will also be referred to as a "similar-temperature area".

Groups of logging devices within the energy grid may be selected in a number of ways, e.g. based on existing information about the energy grid, such as:

1. Based on street name, consumer address coordinates and/or registered GPS coordinates of the respective logging devices.

2. Based on valve closing maps from the district energy grid operator's GIS data.

3. Based on a graph-theoretical representation of the energy grid.

It will be appreciated that the grouping may be based on each of the above information, individually or on a combination of two or even all three types of information. It will further be appreciated that the grouping may be performed in a different manner, including based on other types of information.

The grouping based on street name and/or consumer address coordinates is easy to implement but is also expected to be the most inaccurate. GPS location of the logging devices is sometimes registered by the grid operator upon installation of the logging device. If available, this data may provide valuable additional information for the grouping, thus increasing the accuracy of the grouping.

Some district energy grid operators maintain valve closing maps in their GIS data. Accordingly, if such data is available, it may also provide a useful basis for the grouping of logging devices. Moreover, such maps may give accurate information about the flow direction in the grid, and thus help identifying similar temperature areas. For example, in the example of FIG. 2, the logging devices in each of the sub-areas 2a - 2d may be selected as a respective group representing a similar temperature area, as all service pipes within each group are connected to a linear string of supply line without any branches. In some instances, depending on the grid topology, several of the sub-areas may be combined into a single group. Alternatively or additionally, a subarea may be divided into multiple groups, e.g. if the subarea includes a very long string of supply line.

In order to aid the grouping, a graph theoretical network model can be established from the district energy grid operator's GIS data. Even though the grid operator in many cases does not have detailed pipe dimensions and heat loss coefficients available they often do have a map of their pipes that may be transferred to a graph theoretical model. It is an advantage of the method disclosed herein, that it does not rely on detailed information about the energy grid, such as pipe dimensions, heat loss coefficients of the pipes etc. As can be appreciated from the above, even though access to GIS data or other grid topology data may increase the accuracy of the grouping, and thus the accuracy of the resulting heat loss computation, the grouping of logging devices may even be performed without any topology information about the grid; it may entirely be based on postal addresses of the consumers. In subsequent step S2, the process selects a group of service lines of the district energy grid 1, each selected service line having an associated data logging device 8.

Accordingly, various embodiments of the method disclosed herein perform a determination of energy loss for a thus selected group of logging devices, i.e. for a selected group of service pipes, preferably a group of service pipes connected to a single supply line or otherwise connected to the supply grid such that the supply lines of the selected group receive fluid of approximately the same or at least of similar temperature or where temperature differences between the points of the supply line where the service lines are connected can easily be modelled. Preferably, the selected group of service lines includes between 5 and 50 service lines with associated logging devices, a typical group size is 10 to 30 service lines. Larger groups of service lines increase the statistical accuracy of the computation, but also increase the computational complexity. Larger groups of service lines may increase the risk that the assumption of uniform temperature in the supply line - or the accuracy of a model of temperature variations along the supply line - is inaccurate, thus reducing the accuracy of the temperature loss determination. Generally, when each service line has a logging device associated with it, the selected group of service lines corresponds to a selected group of associated logging devices.

In step S3, the process selects a suitable observation period for which thermal losses of one or more service lines of the selected group of service lines are to be computed. As discussed below, the length of the observation period may be selected in dependence of the rate at which the logging devices record sensor data. Other factors that may influence the choice of observation period may include the size of the selected group of service lines. For example, if the group is small, the accuracy of the estimates of the heat loss may decrease. Increasing the length of the observation period may then increase the accuracy of the estimates. When the observation period is selected to be very long, the assumption of uniform supply line temperature may be less accurate.

In step S4, the process obtains logged sensor data from the logging devices associated with the selected group of service lines. In the following, the number of logging devices in the selected group will be denoted N (A/>=1). The process may obtain the sensor data in a variety of ways. For example, the process may receive the sensor data directly or indirectly from the logging devices. The process may receive sensor data from an entity that transmits the sensor data to the process, or the process may request the sensor data from another entity, or retrieve the sensor data from a suitable data repository.

Various embodiments of the method use sensor data logged at different points in time over a selected observation period of time, e.g. a predetermined observation period, such as over an observation period of between lh and 1 week, e.g. between 6 h and 24 h, e.g. between 12 h and 24 h. In particular, various embodiments use, for each logging device of the selected group, logged sensor data, logged at different points in time over an observation period of time. The duration of the observation period may be selected long enough so that a plurality of data points logged at respective points in time can be collected, i.e. the choice of observation period may depend on the frequency at which the logging devices log measurements. For example, each logging device may log the sensor data at a rate between once every minute and once per day, such as between once every 10 minutes and once every 6h, such as hourly. It will be appreciated that some embodiments, e.g. when the sensor data is logged at a high rate, may accumulate, e.g. average, data over a sampling period and use the thus accumulated data as a logged measurement value. The length of the observation period is preferably chosen short enough such that the temperature in the supply line may be regarded as substantially constant over the observation period. A larger number of data points may increase the statistical accuracy of the determined thermal losses. On the other hand, a larger number of data points also increases the computational complexity of the process. Generally, the use of a large data batch from multiple logging devices and at multiple points in time improves robustness of the heat loss determination. The process may obtain the sensor data in real time from the logging devices or it may retrieve one or more batches of previously logged data for a desired observation period.

In subsequent step S5, the process defines and solves an optimization problem, the optimization problem is formulated such that its solution directly represents the thermal losses of the service lines of the selected group for the selected observation period, in particular without the need for subsequent dynamic modelling of fluid flow through the main grid and without use of measured or computed data pertaining to parts of the energy grid outside the subarea defined by the selected group of service lines. To this end, the process calculates the thermal loss of a service line 5 between a supply line 6 and a logging device 8. The calculation may be based on the supply temperature and the volume and/or flow data logs of one or more logging devices 8.

Various embodiments of the process base the computation on a model of the instant thermal loss of the service line, which may be represented by the following equation:

Q = Cq(T _v - T _m ~) (eq. 1) where C is the volumetric heat capacity of the fluid in the service line, q is the fluid flow in the service line, T _v is the temperature in the supply line from which the service line branches off, and the T _m is the measured fluid temperature of the fluid arriving at the logging device, i.e. T _v represents the fluid temperature at the receiving end of the service line and T _m represents the fluid temperature at the consumer end of the service line.

The logging device may conveniently be a heat meter. Heat meters normally log instant supply and return temperatures across a consumer installation, and/or flow weighted supply and return temperatures. The flow weighted temperature is sometimes referred to as the integrated flow temperature product (ITFP). For the purpose of the present description, the temperature data will simply be referred to either as supply temperature and return temperature or commonly as temperature data. In addition to temperature data, accumulated volume and/or instant fluid flow rate is logged by the heat meter.

Therefore, referring to (eq. 1), the parameter C is a constant known from water properties. Logs of T _m and q can be obtained from the heat meter or another form of logging device, either directly or by suitable pre-processing, depending on the type of sensor data logged by the logging device. Various embodiments of the process described herein estimate Q, or a thermal loss parameter related to Q, based on the logs of T _m and q for a selected group of logging devices and, hence, for a selected group of service lines. Examples of the calculation process will be described in more detail with reference to FIG.

4 below.

In step S6, the process outputs the computed thermal losses or another loss figure of merit indicative of thermal losses of the selected service lines. In some embodiments, the process computes the thermal losses for respective observation periods, e.g. by computing the thermal loss for a new observation period as soon as new logged data is available, e.g. so as to provide estimated recent thermal losses for a sliding window.

Accordingly, optionally, if thermal losses of the one or more service lines of the selected group of service lines are to be computed for further observation periods, e.g. so as to assess how the thermal losses develop over time, the process returns to step S4 where a different observation period is selected.

Similarly, if thermal losses for another group of service lines are to be computed, the process returns to step S3 where a different group of service lines is selected, e.g. so as to allow assessment and/or comparison of multiple subareas of the energy grid.

It will be appreciated that some of the above steps may be omitted or performed in a different order. For example, in some embodiments, the process may process all selected groups for a single observation period before optionally proceeding to another observation period.

FIG. 4 shows a schematic flow diagram of a process of setting up and solving an optimization problem for estimating thermal losses of one or more selected service lines based on the logs of T _m and q for a selected group of logging devices and for a selected observation time.

In initial step S51, the process sets up suitable data structures for representing the optimization problem to be solved and populates the data structures with the fluid temperatures T _m and the associated fluid flow q of the incoming fluid arriving at the logging device at respective points in time within the observation period. To this end, the obtained sensor data may include time-stamped measurements from which the fluid temperature T _m and the fluid flow q of the incoming fluid arriving at the logging device via the associated service line can be derived. In some embodiments, the process may receive the measured temperature T _m and the measured fluid flow q from the corresponding logging device. In other embodiments, the process derives these quantities from the sensor data received from the logging device. For example, in some embodiments, the logging device may only measure the accumulated volume and not the instantaneous flow, or the flow measurements received from the logging device may be inaccurate.

Accordingly, optionally, the process may derive the fluid flow and/or the fluid temperature from the received sensor data. For example, in embodiments where the received sensor data includes time-stamped volume data of the accumulated volume received at the logging device via the service line, the process may calculate the fluid flow as the gradient with respect to time of the accumulated volume using the actual time stamps of the received volume data. The used gradient method is preferably a central difference method, but backward or forward difference or other methods may be used instead.

It will be appreciated that other embodiments may receive other forms of temperature data and/or other forms of data indicative of the fluid flow. Accordingly, in such embodiments, the received sensor data may be pre-processed in a different manner in order to derive the fluid temperature and/or the fluid flow.

The created data structure may be a matrix or another representation suitable for processing the optimization problem. For example, in some embodiments, the process may arrange the data representing the fluid flow pertaining to each logging device in a corresponding time vector or other suitable representation of a time series of data points. For example, this may be done by moving the actual time stamps of the received or derived data points to the nearest time stamp which is and integer multiple of a predetermined sampling time Ts. The logging rate of the logging devices may differ from logging device to logging device or even from sample to sample. The selected sampling time Ts may e.g. be selected so as to match the average rate at which the logging devices log the sensor data. In one embodiment, Ts is selected to be 1 hour, but other sampling times may be used. It will be appreciated that, in other embodiments, the alignment of the data to time stamps may be omitted. Accordingly, some embodiments merely load flow and temperature data pertaining to corresponding points in time within the observation period into one or more suitable data structures. Accordingly, the process is based on batches of sensor data, each batch including measurement data from one or more logging devices associated with a given measurement time. The number of batches will be denoted s. Preferably, there are a plurality of batches (s > 1), i.e. the process uses sensor data associated with multiple different points in time within an observation period.

Optionally, the process processes the created time vector such as to remove equal time stamps for the same logging device.

In step S52, for a data period of s sampling periods, the process forms a data matrix X and an input vector y that belong to a model of the form y = XP (eq. 2)

Here, P = [P _1; P ₂ , ... , P _N ] ^T is a parameter vector where each parameter Pi relates to the thermal loss of a respective service line /.

Accordingly, the model is represented as a system of equations, linear in the thermal loss parameters Pi, which are the unknowns of the system.

The parameters in P may, for example, relate directly to losses Q in [W] (Watt), or to losses in [W/m] (Watt/meter), or to losses in [W/m/k] (Watt/meter/Kelvin) or to losses in [W/K] (Watt/Kelvin). Accordingly, the thermal loss parameters in P generally represent suitable thermal-loss-related figures of merit.

As an alternative, the parameter vector P may, in addition to the P _± to P _N service line loss related parameters, have one or more other parameters not relating to service line losses. As will be described in more detail below, such one or more additional parameters may for example contain a temperature parameter such as T _v in (eq.l).

In step S53, the process numerically solves the model for the parameter vector P, which is returned as a result of the process.

The exact form of the data matrix X, the input vector y and the parameter vector P may differ for different embodiments and depend on the exact model used.

In a first embodiment, the model used is directly derived from (eq.l) and may be written in the form y = XP by letting y be dependent on the temperature T _v in the supply line: where and k is the batch number and s is the number of batches. Therefore, k may be advanced with the chosen time step, and the data batches may be stacked to form y(T _v ~) and X until k reaches a predefined limit s. For example, if batches are collected each hour and if an 8 hour observation time window is used then s = 8. N is the number of selected logging devices that have been active during data collection of the s batches. It will be appreciated that the different batches do not need to be uniformly spaced in time, nor do the sensor data pertaining to different logging devices be equally spaced in time. Various embodiments of the method disclosed herein perform a statistical analysis of the sensor data pertaining to different logging devices and different points in time rather than attempting a fluid-dynamic modelling of the energy grid.

In most instances, in particular where s > 1, the above system of equations is overdetermined, i.e. there are more equations than unknowns. While such an overdetermined system in general does not have an exact solution, an approximate solution may be determined as a solution of an optimization problem, in particular by means of the method of ordinary least squares. For the system of (eqn. 2) the least squares solution may be obtained by solving the optimization problem

The solution of this optimization problem can be written as

P = X ^T X)~ ¹ X ^T y where the superscript T indicates a matrix transpose. The above is an approximate solution when no exact solution exists, while it is an exact solution when one does exist. In the present case, the input vector depends on the fluid temperature T _v in the supply line, which is normally unknown. In the present embodiment, the temperature T _v is assumed to be equal for all service lines of the selected group. Accordingly, as described above, the group of logging devices is preferably selected such that this assumption is at least approximately fulfilled. Accordingly, an approximate solution may be found by formulating a constrained optimization problem for T _v .

To this end, when the s data batches have been collected, the model in (eq.2) may be solved by minimizing the squared error under the constraints that the losses are positive, that is

T _v = arg min Z(y (T _v ) - XP) (eq. 3) l _v subject to: P = X ^T X)~ ¹ X ^T y(T _v ) > 0 The resulting least-square solution P represents the estimated heat losses of the set of service lines during the observation time window. After calculation of the parameter vector P, a new data batch may be collected. The solution to (eq.3) can be found by known numerical optimization algorithms.

In another embodiment, the same model as above may be solved as an unconstrained least squares problem by moving the unknown temperature T _v in the supply line into the parameter vector. The formulation of the problem is then where and the solution to the parameter vector is calculated directly

P = X ^T X)- ¹ X ^T y

In yet another embodiment of the process, the loss calculation in (eq.l) can be extended by adding some structure of the loss of the service line, e.g. according to the following modified model:

A loss proportional to the arithmetic mean temperature difference between the service line fluid and the surrounding temperature T _g is assumed in (eq.5). For a service line, the surrounding temperature would be the ground temperature. To solve the model, T _g is now a needed input. It may be found via an outdoor temperature log and an associated model for the ground temperature, or by setting it as a predetermined constant, or by letting it be solved by the optimizer as shown below. The model can be written on the form y(T _v ) = X(T _v , T _g )P (eq.6) where and

The parameter vector P for this model formulation now contains loss parameters B _t with the unit [W/K], A solution may be found similar to the constrained formulation of the first embodiment above, only now T _g is also a parameter to be found by the optimizer, i.e.

In the above constrained solutions according to (eq.3) and (eq.7), no constraints are set on T _v and/or T _g , but due to numerical robustness it may be beneficial to constraint these parameters as well. That is, if it is known that the considered district energy system is for district heating, a constraint may be incorporated on T _v as well, e.g. adding to the solution in (eq.7) subject to:

V /c e {1, ...,

It will be appreciated that the above embodiments do not require that all logging devices of the selected group of logging devices, e.g. all heat meters in a "similar temperature area," deliver sensor data at all time stamps. Accordingly, the process may not be able to determine the flow in the supply line feeding into the selected group of service lines. However, in some embodiments, the selected group of logging devices may include a logging device associated with each service line branching off a particular selected branch of the supply line, e.g. all logging devices of subarea 2b in the example of FIG. 2. In such an embodiment, if all the selected logging devices deliver flow data at all time stamps, and if there is no, or at least no significant, flow by-pass in the branch, the process can determine the flow in the section of supply line from which the selected service lines branch off. This information can be exploited by the process to estimate the condition of that particular section of the supply line as well, e.g. by setting up a model based on the temperature decay in the supply line and the loss of the service line, e.g. as follows:

Q = Cq(T _v - T _m ~) (eq.l) where (eq.8) is a Taylor approximation of a model of an exponential temperature decay along the length I of a section of the supply line and q _v is the flow in that section of the supply line. Using this information, a model can be set up as follows: y(a, T _fl ) = X(a)P (eq.9) where and

Here, a is the loss parameter of the supply line and q _t j is the flow in the supply line between consumers i and j. This flow can be calculated exactly, if all logging devices deliver data at all time stamps, and there are no unknown by-pass flow between the supply line and the return path. is the length of the supply line between consumers i ad j. The model formulation is usable if these lengths are known. Still, in comparison to prior art methods that depend on further details of the energy grid, such as pipe diameters, insulation class and heat loss coefficients, the present embodiment estimates conditions of the supply line and the service pipes directly. solution to (eq.9) may be found in a similar manner as for the previous embodiments, e.g. as

{a, T _g } = arg (eq. 10) subject to:

It will be appreciated that the supply line loss model used in (eq.8) may be combined with the service line loss model used in (eq.5) to an optimization problem with a similar structure as (eq.9), i.e. having outer variables adhering to line loss models that are optimized and the service line losses are calculated as constraints as in (eq.3), (eq.7) or (eq.10). In case the service line lengths are known, the loss parameters calculated in (eq.3), (eq.7) or (eq.10) can easily be converted to a per-length figure of merit. This may give a more easily comparable figure when comparing service line losses between consumers as such pipes are sometimes of very different lengths within the same "similar temperature" area. The thermal loss of a service line in new condition is also normally given as a per-length figure of merit. However, as mention earlier, especially for old service lines, the service line lengths of are often not available. Still, embodiments of the process disclosed herein can assist the energy grid operator in identifying the service lines that have the highest losses, which is still very usable information for the district energy grid operator.

Accordingly, in each of the various embodiments described above, a method for estimating heat loss parameters of a district energy grid is disclosed where the estimated heat loss parameters are indicative of, in particular directly related to, thermal losses of service lines. Some embodiments of the method comprise the steps of: selecting a group of N logging devices, in particular logging devices within a "similar temperature area;" for example, the boundaries of the "similar temperature area" may be selected based on a street name, GIS data describing the district energy network, knowledge of valve locations in the district energy network, or based on other suitable information. reading time-stamped sensor data logs from the N logging devices (/V>=1) into a data structure, deriving N flow rates associated with the N logging devices from the received sensor data logs, constructing a time vector for each logging device; this may e.g. be done by moving the actual time stamps of the logging device to the nearest time stamp which is and integer multiple of a sampling time T _s . T _s is preferably 1 hour or less, optionally processing the constructed time vector so there are no equal time stamps, for an observation time, e.g. an observation time of s sampling times, e.g. of s hours, constructing a data matrix X and an input vector y that belong to a model which is linear in the thermal loss parameters, such as y = XP, where the 1 parameter vector P = [P _1; P ₂ , ... , PN] ^T has N thermal loss parameters that relate to respective thermal losses of N service lines, numerically solving the constructed model for the parameter vector P.

The data matrix and the input vector preferably contain data for multiple time stamps of the constructed time vector.

Numerically solving the constructed model may comprise solving a constrained optimization problem. Solving the constrained optimization model may involve minimizing or maximizing a cost function over at least one additional parameter that is associated with a temperature and/or a loss parameter that is associated with a supply line. Alternatively or additionally, solving the constrained optimization model may involve minimizing or maximizing a cost function over at least one additional parameter that is associated with a ground temperature.

FIGs. 5A-F illustrate examples of measured fluid temperatures and flows of respective service lines, and examples of computed results obtained by an embodiment of the method described herein. In this example, the selected group included four logging devices, labelled "Meter 1", ..., "Meter 4", respectively. It will be appreciated that other examples of selected groups of logging devices may include a different number of devices.

In particular, FIG. 5A shows logged temperature data for a selected group of logging devices and the estimated temperature 12 in the supply line, which has been computed by the method disclosed herein. The computed supply line temperature is denoted "Estimated Main".

FIG. 5B shows the logged flow data from the selected group of logging devices.

FIGs. 5C though 5F show resulting values of an estimated heat loss parameter for the respective service lines associated with the respective logging devices Meter 1 through Meter 4, respectively, including indications of +/- 3 standard deviations. As can be seen from FIG.s 5A-B, one of the logging devices ("Meter 4") has registered a relatively high flow 11 in the associated service line, while the corresponding temperature 10 registered by said logging device was relatively low. This indicates a high thermal loss of the corresponding service line which embodiments of the method disclosed herein help to identify. Accordingly, a comparison of FIG. 5F with FIG.s 5C-E shows that the service line associated with Meter 4 has an unusually high heat loss.

It is generally useful for an energy grid operator to identify service lines having a high thermal loss, as it may be useful to eliminate the cause of abnormally high thermal losses. Possible causes of high identified thermal losses may include defective or incorrectly mounted logging devices, insufficient or damaged thermal insulation, pipe leakages, etc. In the example of FIG. 5A-F, all service lines of the selected group had similar lengths (in this particular case about 15 m), so the computed thermal losses were directly comparable without converting them into specific losses per unit length. In the present example, a subsequent inspection of the service line corresponding to Meter 4 revealed that the service line had a leak and the pipe insulation was wet.

The various embodiments described herein are applicable to thermal energy carrying pipe networks where a group of consumers are supplied from a common supply line, and where a temperature log of the flow out of the service line connecting the consumer to the supply line is available together with a log of the amount of medium flowing in the service line. Examples of data logs relating to the amount of medium flowing in the service line are logs of the accumulated volume or logs of the instantaneous flow rate. In such a case the thermal pipe loss or a thermal pipe loss related coefficient of the service line can be estimated by the various embodiments disclosed herein.

Generally, embodiments disclosed herein may be applied to energy grids of different types and sizes, including to energy grids located inside or outside of buildings. It will be appreciated that the application of the various embodiments typically is most useful in respect of energy grids where visual pipe inspection is difficult. Therefore, a district energy grid with automated data logging at the consumers and where the pipes are buried in the ground is a preferred application for embodiments of the invention disclosed herein.

In summary, various embodiments of the aspects disclosed herein may be summaries as follows:

Embodiment 1: A computer-implemented method for determining thermal loss in a district energy grid, the district energy grid comprising one or more supply lines and a plurality of service lines for transporting a fluid to a plurality of consumers, each supply line feeding the fluid into a plurality of service lines and each service line connecting a consumer with one of the one or more supply lines, the method comprising:

- obtaining measured sensor data from one or more sensors associated with one or more selected service lines of the plurality of service lines;

- determining, from the received sensor data, respective observed fluid temperatures and observed fluid flows of the transported fluid received at respective points in time via respective ones of said one or more selected service lines;

- obtaining a set of equations, the set of equations comprising a plurality of equations relating the observed fluid temperatures and the observed fluid flows with thermal loss parameters indicative of thermal losses of respective ones of said selected one or more service lines, and

- computing a solution of the set of equations and computing a result value of at least a first thermal loss parameter indicative of a thermal loss of a first service line of said selected one or more service lines from the computed solution of the set of equations.

Embodiment 2: A method according to embodiment 1, wherein the thermal loss parameters represent unknown variables of the set of equations.

Embodiment 3: A method according to any one of the preceding embodiments, wherein the plurality of equations are linear in the thermal loss parameters. Embodiment 4: A method according to any one of the preceding embodiments, wherein the set of equations is an overdetermined set of equations and wherein the solution of the set of equations is an approximate solution of the overdetermined set of equations.

Embodiment 5: A method according to embodiment 4, wherein computing the solution of the set of equations comprises computing the solution as a solution to a constrained optimization problem.

Embodiment 6: A method according to embodiment 4 or 5, wherein computing the solution of the set of equations comprises applying a least-square method for determining the approximate solution of the overdetermined set of equations.

Embodiment 7: A method according to embodiment 5 or 6, wherein computing the solution of the set of equations comprises minimizing or maximizing a cost function over at least one additional temperature parameter indicative of a temperature and/or a temperature loss associated with the supply line.

Embodiment 8: A method according to any one of embodiments 5 through 7 , wherein computing the solution of the set of equations comprises minimizing or maximizing a cost function over at least one ground temperature parameter that is associated with a ground temperature.

Embodiment 9: A method according to any one of the preceding embodiments, wherein the one or more selected service lines include a plurality of service lines, and wherein each of the plurality of equations relates a thermal loss parameter of a respective service line with an observed fluid temperature and an observed fluid flow associated with said respective service line.

Embodiment 10: A method according to any one of the preceding embodiments, wherein the set of equations comprises a first equation relating an observed fluid temperature and an observed fluid flow of the first service line at a first point in time with the first thermal loss parameter. Embodiment 11: A method according to embodiment 10, wherein the plurality of equations further comprises another equation associated with a second point in time and relating an observed fluid temperature and an observed fluid flow of the first service line at said second point in time with said first thermal loss parameter, the second point in time being different from the first point in time.

Embodiment 12: A method according to embodiment 10 or 11, wherein the one or more selected service lines include a first and a second service line, different from the first service line, and wherein the plurality of equations comprises another equation associated with the second service line and relating an observed fluid temperature and an observed fluid flow of said second service line with a second thermal loss parameter indicative of a thermal loss of said second service line.

Embodiment 13: A method according to embodiment 12, wherein the first equation and the another equation associated with the second service line each include a temperature parameter associated with the supply line, the temperature parameter representing an unknown variable of the first equation and of the another equation associated with the second service line.

Embodiment 14: A method according to any one of the preceding embodiments, wherein the one or more selected service lines include a plurality of service lines all connected to a single supply line.

Embodiment 15: A data processing system configured to perform the steps of the method defined in any one of embodiments 1 through 14.

Embodiment 16: A district heating or cooling system comprising:

- a district energy grid, the district energy grid comprising one or more supply lines and a plurality of service lines for transporting a fluid to a plurality of consumers, each supply line feeding fluid into a plurality of service lines and each service line connecting a consumer with one of the one or more supply lines, - a plurality of sensors configured to output sensor data indicative of measured fluid temperatures and associated fluid flow of the transported fluid received at respective consumers via the respective service lines,

- a data processing system according to embodiment 15.

Embodiment 17: A computer program comprising computer program code configured, when executed by a data processing system, to cause the data processing system to perform the steps of the method according to any one of embodiments 1 through 14.

Embodiments of the method described herein may be computer-implemented. In particular, embodiments of the method may be implemented by means of hardware comprising several distinct elements, and/or at least in part by means of a suitably programmed data processing system. In the apparatus claims enumerating several means, several of these means can be embodied by one and the same element, component or item of hardware. The mere fact that certain measures are recited in mutually different dependent claims or described in different embodiments does not indicate that a combination of these measures cannot be used to advantage.

The term "obtaining a set of equations" as used in this disclosure, refers to the computer system automatically or semi-automatically, setting up the set of equations according to operational parameters of the district energy grid. For example, the computer system may take into account the size or number of users of the district energy grid, input data for one or more individual zones of the district energy grid, and/or any other relevant input parameters relating to the district energy grid, where the input data and/or other relevant input parameters may be obtained automatically and/or through user input.

It should be emphasized that the term "comprises/comprising" when used in this specification is taken to specify the presence of stated features, elements, steps or components but does not preclude the presence or addition of one or more other features, elements, steps, components or groups thereof.