GB2448896A

GB2448896A - Energy management system

Info

Publication number: GB2448896A
Application number: GB0708448A
Authority: GB
Inventors: Peter Boait
Original assignee: De Montfort University
Current assignee: De Montfort University
Priority date: 2007-05-02
Filing date: 2007-05-02
Publication date: 2008-11-05
Anticipated expiration: 2027-05-02
Also published as: GB0708448D0; GB2448896B; GB2448896A8

Abstract

The present invention provides an energy management system for automatic management of one or more energy conversion devices 5, 8 located in a domestic, commercial or industrial building. The system includes a plurality of sensors 16, 18 measuring parameters including temperature and electrical load and providing energy sensor data. Means are provided for automatic modeling of energy conversion devices, thermal modeling means for modeling thermal properties of a building, occupant behavior recognition means, generating a predictive model of the behavior patterns of the occupants, temperature tolerance identification mean, and energy demand prediction means for predicting energy demands, all provided within a single electronic management unit 10. The electronic management unit 10 uses the information to schedule the operation of the energy conversion devices 5, 8 in an manner meeting the predicted energy demands.

Description

TITLE

Energy management system

DESCRIPTION

Technical field

The present invention relates to the field of household and building energy management, and in particular to the efficient use of energy from different sources.

The apparatus and methods require little or no user intervention because thcy are self-configuring in that they automatically acquire as much as possible of the information needed for efficient operation, and make safe assumptions if information is not available at any time.

Background to the invention

1'he occupants of a household or building in a temperate climate typically need energy in three forms, electricity for lighting and electrical appliances, space heating of occupied rooms, and hot water for washing. The electricity is conventionally obtained directly from a connection to the local mains electricity supply, while space heating and hot water heating is delivered by the conversion of a primary energy source such as gas or electricity into heat by one or more energy conversion devices such as a gas boiler or an immersion heater, for example. These energy conversion devices are normally regulated by a thermostat control so that a certain temperature (setpoint) that is determined by the occupants is achieved in rooms and for hot water.

In other words, there is a room temperature setpoint that represents a desired temperature for space heating and a hot water temperature setpoint that represents a desired temperature for hot water. The times when space heating or hot water heating is desired are also entered manually into simple mechanical or electronic control units.

There are certain inefficiencies associated with this conventional arrangement.

Manual setting of times when space heating or hot water heating is needed is often not performed accurately. Also, temperature setpoints are not adjusted to reflect reduced heating needs when weather improves. The small physical size of a typical mechanical or electronic control unit makes its manual controls difficult to use particularly for the elderly or disabled. Where more than one energy conversion device is available to meet an energy need, their relative efficiencies are not exploited.

For example, in many homes the water in the hot water tank can be heated either by an immersion heater or by a gas boiler. When factors such as the amount of hot water that is required, the respective efficiencies of the energy conversion devices and the prevailing energy tariffs are taken into account it may preferable to use the gas boiler during the winter but cheaper and more efficient in summer to heat the water using the immersion heater, for example. Few consumers are able to make this determination accurately and then manually switch between the gas boiler and the immersion heater on an optimum basis.

The advent of climate change makes these limitations much more significant.

Weather is becoming more variable and extreme, leading to more frequent and larger changes in external ambient temperature from one day to the next. This increases the inefficiency arising from inaccurate setpoint and time settings. Higher summer temperatures will lead to more frequent adoption of air conditioning, thereby introducing another energy conversion device into the household or office. To reduce carbon emissions, consumers and businesses are installing new types of energy conversion devices (renewable energy devices) such as solar water heaters, solar photovoltaic generators, wind turbines and combined heat and power (CHP) generators which generate electricity as a by-product when heat is required. These renewable energy devices each have their own limitations and requirements for optimum use making the task of managing the total set of energy conversion devices within a household or building much more complex.

There is therefore a need for an energy management system that is suitable for domestic and small office or industrial buildings and which can automatically sense the energy needs of the occupants and deploy to optimum effect whatever energy conversion devices are available. It should preferably be able to recognise, characterise and manage any energy conversion device without the need for a complex installation process. Such an energy management system would reduce the cost of installing an energy conversion device (and in particular those devices that are intended to reduce carbon emissions) and make sure that their benefits are fully obtained.

Summary of the invention

The present invention seeks to overcome the above-mentioned problems and provides an energy management system for efficient and automatic management of one or more energy conversion devices located in a building, the system comprising: a plurality of sensors measuring energy related parameters such as temperature and electrical load and providing energy sensor data; device recognition means for automatic recognition, characterisation and modelling of the one or more energy conversion devices from the energy sensor data; thermal modelling means for automatic characterisation and modelling of the thermal properties of the building from the energy sensor data; occupant behaviour recognition means for automatic recognition of whether or not the building is occupied at any particular instant in time and for generating a predictive model of the behaviour patterns of the occupant or occupants of the building; temperature tolerance identification means for identifying the temperature tolerances of the occupant or occupants of the building from their response or absence of response to temperature setpoints that are varied automatically; energy demand prediction means for predicting energy demands depending on the weather, building occupancy and behaviour patterns of the occupant or occupants of the building; and energy management means to use the information obtained by the device recognition means, the thermal modelling means, the occupant behaviour recognition means, the temperature tolerance identification means and the energy demand prediction means to schedule the operation of the one or more energy conversion devices in an efficient manner thereby meeting the predicted energy demands.

The present invention further provides a method of energy management for efficient and automatic management of one or more energy conversion devices located in a building, the method comprising the steps of: measuring at least one energy related parameter (such as temperature and electrical load, for example) to obtain energy sensor data; automatically recognising, characterising and modelling the one or more energy conversion devices from the energy sensor data; automatically characterising and modelling the thermal properties of the building from the energy sensor data; automatically recognising whether or not the building is occupied at any particular instant in time and generating a predictive model of the behaviour patterns of the occupant or occupants of the building; identifying the temperature tolerances of the occupant or occupants of the building from their response or absence of response to temperature setpoints that are varied automatically; predicting energy demands depending on the weather, building occupancy and behaviour patterns of the occupant or occupants of the building; and scheduling the operation of the one or more energy conversion devices in an efficient manner to meet the predicted energy demands.

For the purposes of the following description, the term "building" will be taken to include any domestic, commercial or industrial building including (but not limited to) a private house, flat, apartment or bungalow, a shop, warehouse, office building or factory.

The energy management system effectively integrates all of the automatic information gathering, prediction and optimisation methods relating to building energy management that are already used individually or in limited combinations in a single information processing and control apparatus that requires a minimum amount of user

input. In summary these known methods are:

* automatic recognition and characterisation of energy sources (i.e. renewable energy sources such as solar radiation, air and ground heat, wind, waves and tides and controllable auxiliary energy sources such as mains electricity and mains gas) and energy conversion devices, including interdependencies between types of energy source as occurs for example with combined heat and power (CUP) generators and heat pumps; * eflicient use of energy conversion devices with an energy storage capability such as electrically heated thermal storage radiators; * automatic characterisation and modelling of the thermal properties of a building, particularly thermal capacity, thermal load, solar gain and heat loss due to wind; * recognition of building occupancy from energy sensors (for example sensors detecting electrical load and hot water usage) and predictive modelling of the behaviour of the occupant or occupants of the building; * identification of room temperature preferences and tolerances of the occupant or occupants of the building from their response to an automatically time varying room temperature setpoint; * prediction of energy demands dependent on the weather, household occupancy and behaviour patterns, building thermal properties and room temperature preferences; * applying the available energy sources and energy conversion devices in the most efficient way that will meet the predicted energy demands by controlling according to an optimum schedule those energy sources and energy demands that are amenable to control; and * employment of default assumptions and simple control methods which are progressively improved and made more efficient as information is collected and predictions made more accurate.

While these methods are known in general terms, their integration into a single energy management system (practically embodied in a single information processing and control apparatus) allows a lower cost practical realisation because the information gathered for one method is also employed by others. The benefits of convenience in installation and use, and in energy efficiency, are also greater than the sum of that offered by the individual methods. Moreover, to achieve the necessary integration of these methods, each must be subject to adaptations that are specific to the present invention.

To gather the necessary information, the energy management system preferably includes a plurality of sensors which take measurements of the energy flows and temperatures associated with the building. As a minimum the sensors will take measurements of the external air temperature, internal air temperature, stored hot water temperature, electricity consumption, import and export of electricity from the mains supply and consumption of any chemical energy source such as mains gas, heating oil or woodchips etc. The energy management means may use initial values and control models that are progressively improved as more energy sensor data is accumulated over time from the energy sensors so that energy demands can be predicted more accurately and the operation of the one or more energy conversion devices can be scheduled in a more efficient manner. The energy management system may also allow the occupant or occupants of the building to enter user data on initial installation and exceptionally thereafter. The user data may be used to determine the initial values andlor control models that are then progressively improved as more energy sensor data is accumulated over time.

The device recognition means preferably uses the energy sensor data to derive a profile of the changes over time in the energy related parameter, correlates the derived profile with one or more reference profiles, each reference profile corresponding to a particular energy source, and uses the results of the correlation to identify the type of energy conversion device.

To control the one or more energy conversion devices (or where applicable, electrical appliances) the energy management system is able to issue commands requesting that: * the energy conversion device that is the primary source of space heating be activated or stopped; * the energy conversion device that is the secondary source of space heating be activated or stopped; * the energy conversion device that is the primary source of room cooling be activated or stopped; * the energy conversion device that is the secondary source of room cooling be activated or stopped; * the energy conversion device that is the primary source of hot water heating be activated or stopped; * the energy conversion device that is the secondary source of hot water heating be activated or stopped; * any electrical appliance whose operation can be controlled be activated or stopped; * any energy conversion device capable of storing energy start or stop storage.

The allocation of primary and secondary commands to particular energy conversion devices is performed by the installation technician on the basis of an a priori assessment of which has lower cost in operation. For example, the primary command to the primary source of hot water heating might be directed at a gas boiler since that is likely to have the lower cost in the majority of circumstances, and the secondary command to the secondary source of hot water heating might be directed to an electric immersion heater. The exact nature of the energy conversion device assigned to respond to each command is not important since its characteristics will be automatically discovered, although for some applications of the present invention it may be convenient to specify that a particular command be used for a class or type of energy conversion device. I t is also not essential that all of the commands are implemented or used in a particular embodiment of the present invention. For more complex embodiment of the present invention it would be possible to allocate tertiary or further additional commands on the same basis.

Energy conversion devices that respond to these commands are referred to as being "under control" while those which are recognised and modelled by the energy management system and contribute to the energy needs of the building, but are not amenable to direct control, are referred to as "under supervision". Electrical appliances such as washing machines, dishwashers and tumble dryers, for example, can be controlled to operate when the cost of electricity is determined to be low.

Automatic recognition and characterisation of energy conversion devices A suitable method for automatic recognition and characterisation of renewable energy devices is described in GB 0702392.2 to the present applicant. The method starts with the collection of energy data from energy sensors which measure energy flows such as the electricity consumption of a building and the rising temperature in a hot water tank. These energy flows can be used to prepare a profile of the variations in energy flow over a time period such as 24 hours (i.e. a daily profile). The profile can be correlated against a set of stored reference profiles to identify the type of energy conversion device that is causing the energy flow. For example, if the profile of the variations in energy flow is considered to match a reference profile that corresponds to solar radiation intensity then the electronic control unit will identify the energy conversion device as a solar water heater. The correlation (which may take place in one or both of the time and frequency domains) is preferably performed using known mathematical techniques for correlation and pattern recognition that give a quantitative assessment (referred to in this description as the "correlation factor") of the accuracy of the match when the profile is compared against a reference profile.

Once recognition has been performed then it is possible to generate a model of the behaviour of the energy conversion device that can be used to predict its performance and energy efficiency at a future time taking into account relevant parameters such as the expected weather conditions and the thermal characteristics of the building.

In the energy management system of the present application, the method is extended to be capable of recognising, and generating a predictive model for, any kind of energy conversion device that might reasonably be employed under the control or supervision of the system. The recognition function may not be needed where a command from the energy management system is by prior design allocated to a specific class or type of energy conversion device, hut the predictive model is always required. The important parameters produced by the predictive model are the response times of the energy conversion device (i.e. how long it takes to start and stop a desired output such as space heating from the time the command is given), the energy efficiency (i.e. how much of the desired output such as space heating is obtained from the energy source inputs such as mains electricity and mains gas), and the maximum desired energy output of which the energy conversion device is capable.

Table 1 shows how different energy conversion devices may be identified and modelled from recognition of patterns in the measurements from particular sensors, and the presence or absence of correlation between the measurements and the issue of commands. The contents of Table 1 are not exhaustive but represent a general cross-section of possible sensor measurements and energy conversion devices that might be encountered in a domestic, commercial or industrial building.

Sensor measurements Patterns observed Energy conversion device identified and modelled Hot water tank temperature, Rising water temperature on Electric immersion heater electrical load command correlated with electrical load in range I -10kW Hot water tank temperature Rising water temperature Solar water heater correlated with reference profile for solar radiation intensity, no correlation with command Hot water tank temperature, Rising water temperature on Gas boiler with electric water electrical load, gas command correlated with gas pump consumption consumption and small electrical load Hot water tank temperature, Rising water temperature on Gas combined heat and power electrical load, metered command correlated with gas (CHP) generator with electric electrical power, gas consumption, small electrical load, water pump consumption and difference between electrical load and metered power internal air temperature, Heating effect on command Gas boiler with electric water electrical load, gas correlated with small electrical pump consumption load and gas consumption Internal air temperature, Heating effect on command Heat extraction by fan from electrical load, external air correlated with small electrical storage in a thermal mass temperature load, heat output decays with time Internal air temperature, Cooling effect on command Air conditioning unit electrical load, external air correlated with electrical load, temperature cooling output and electrical load dependent on external temperature.

Internal air temperature, Heating effect on command Heat pump electrical load, external air correlated with electrical load, temperature heating output and electrical load dependent on external temperature Sensor measurements Patterns observed Energy conversion device identified and modelled Ekctrical load, electrical Detection of locally generated Micro wind turbine power imported from mains electricity with pattern matching profile for wind turbine capacity factor Electrical load, electrical Detection of locally generated Solar photovoltaic generator power imported from mains electricity with pattern matching profile for solar radiation intensity

Table I

Efficient use of energy conversion devices with an energy storgg capability The most common example of an energy conversion device with energy storage capability is the electrically heated thermal storage radiator. This uses off-peak electricity (known as Economy 7" from the 7 hour overnight period in which it is available) to heat ceramic bricks with a high thermal capacity in an insulated cabinet.

This heat may then be released later for space heating, for example by circulating air through the bricks with a fan or by natural convection. The energy management system of the present invention makes more efficient use of these storage radiators through the application of two methods. The first is automatic characterisation and modelling of the storage properties using the same techniques as described above for other energy conversion devices. The key properties to be modelled are the amount of energy that can be stored, the rate at which it can stored, and the rate of "leakage", in other words the unintended loss of stored heat to the surroundings due to imperfect insulation. I'he properties concerned with extraction of heat from the storage radiator will be determined through its response to primary or secondary room heating commands.

The second method is to ensure that the least cost amount of energy is stored within the storage radiator to meet expected future heating requirements. This will form part of the computation of a least cost schedule as further described below, which will take account of the modelled storage properties from the first method.

This aspect of the present invention may be used with other forms of energy storage device such as an electrical battery, which might be used to store electricity from an intermittent renewable source such as a solar photovoltaic generator, for example. In this example the energy management system would automatically discover the capacity of the battery and its efficiency (i.e. the ratio of energy output to input) and schedule the most efficient amount of electricity to be stored when there is a surplus available from the solar photovoltaic generator, relative to use of the surplus for other controllable purposes such as space heating or air conditioning (space cooling).

A further example of an energy storage device which the present invention may be employed to manage is a controllable air vent or fan, which by regulating air flow causes heat to be stored by, or released from, a thermally massive structure within or forming part of the building.

Automatic determination and modelling of the thermal properties of a building The automatic determination of the key thermal properties of a building is known for example from GB 2218540 and GB 2212949. These describe a control unit for a central heating device that is configured with the heat output of the device so it can maintain records of the amount of heat that has been put into a building. It also has temperature sensors giving readings of internal and external temperature so that it can relate the heat input to the temperature measurements to calculate the amount of heat that has to be provided to maintain an internal setpoint with a given external temperature. The specific heat loss of the building is then given by the heat input divided by the difference between internal and external temperatures. Once this figure is obtained then the specific thermal capacity of the building can be determined from the rate of temperature loss overnight when no heating is being provided; this being approximately equal to the specific heat loss multiplied by the overnight difference between internal and external temperatures multiplied by the time taken for the internal temperature to fall by one degree. -13-

The energy management system of the present invention can make use pf this method but does not need prior knowledge of the heat output of the heating device because this is characterised automatically as described above. The method is also augmented in the energy management system by measuring and predicting electrical load so that the energy dissipated in the building from electrical appliances is known and predicted as well that from heating devices. This provides more accurate estimates of specific heat loss and specific thermal capacity. Then if the heat loss rate on a given day is higher than expected given the external air temperature, the energy management system can determine that either high wind speeds are occurring, or there is an unusual exposure to air infiltration, for example due to a window left open.

Where a means to obtain a weather prediction is available (for example from a radio broadcast or barometric or other sensor) the energy management system can discriminate between these possibilities. If an exceptional "window open" condition is identified, it can cut the heating output to avoid energy wastage. In other words, the operation of the one or more energy conversion device can be controlled in response to the detection of an abnormal thermal load.

l'he thermal properties of the hot water tank are also characterised by the energy management system using the same method as for the building. The heat input and temperature variations are measured allowing the heat loss rate and thermal capacity to be determined. The thermal capacity also provides the volume of the hot water tank since the specific heat of water is known. Poor insulation of the hot water tank can therefore be automatically identified.

These thermal parameters for the building and its hot water tank allow simple models to be constructed from which heat demands and temperatures can be calculated under different environmental and occupation scenarios.

Recognition of building occupancy and predictive modelling of behaviour -14-Recognition of building occupancy from energy sensors and predictive modelling of the behaviour of the occupants is known from GB 2408592 and GB 0522544.6. GB 2408592 describes how monitoring electrical load can be used to detect occupation of a building and determine the behaviour patterns of the occupant or occupants of the building such as the time they usually get up in the morning and the days and times when they go out to work. GB 0522544.6 to the present applicant describes a practical method of constructing a behaviour model from lifestyle and comfort settings which are initial user inputs followed by Bayesian inferencing to refine and evolve the model. The user may select one of a predetermined number of different lifestyle settings that best fits the occupant's typical pattern of behaviour. A list of possible settings might include: * The occupant is in the building most of the time.

* The occupant is out at work from Monday to Friday and in the building on Saturday and Sunday.

* The occupant is out of the building at different times depending on his or her shift pattern at work.

* The occupant is in and out of the building on an irregular basis.

The user can also select if the method should operate in a comfort mode or an economy mode. If the user selects the comfort mode then the decision making processes will be biased in favour of providing or maintaining the supply of space heating. However, if the user selects the economy mode then the decision making processes will be biased in favour of delaying or ending the supply of space heating so that the system is operated in a more cost-effective and environmentally-friendly manner.

GB 0522544.6 also shows how a sensor fusion technique can be used to improve the accuracy of the behaviour model by combining data from more than one sensor, for example by combining data from hot water tank temperature sensors and electrical load sensors. Other non-energy related sensors can also be used to give an indication of activity within the building. For example, one or more activity sensors can be -15 positioned at any suitable locations within the building and can use any known technology such as microwave, infra-red or acoustic to determine if the occupant or occupants are active and provide an output signal. Data from the activity sensor can be combined with the data from the hot water tank temperatures and electrical load etc. using the sensor fusion technique.

These methods are extended in the energy management system of the present invention by using the predictions of the behavioural model, whichare derived from electrical load, to predict future electrical load. This allows the heating effect of electrical appliances (which in some cases can be relatively large) to be taken into account in scheduling the operation of devices performing space heating or cooling.

Identification of air and water temperature preferences and tolerances In order to be able to schedule the operation of a space heating or cooling device such as an air conditioning unit at the optimum time it is necessary to make use of the thermal capacity of the building which acts as an energy store. The capacity of this store is proportionate to the air temperature tolerance of the occupant or occupants of the building. For example, if they are comfortable within a temperature range of 20- 22 C then about twice as much thermal energy storage is available compared to a building where the occupant or occupants insist on temperatures between 20-21 C.

The greater the storage capacity the longer the time window over which operation of a space heating or cooling device may be brought forward or delayed in time.

It is therefore important to this energy management system to find out the temperature tolerance of the occupant or occupants of the building. This is preferably performed with an air temperature regulation process for which the room temperature setpoint varies during the course of the day. Energy management systems which make use of a time-varying air temperature setpoint are known, for example from EP 0444308 and US 2003/0050737. These methods require the user to enter a desired time-varying temperature profile whereas for the energy management system of the present invention a time-varying profile is executed automatically based on the lifestyle, comfort expectation, and behaviour knowledge gained using the method previously -16-described. The user then indicates when the temperature provided by the automatic time-varying temperature profile is unacceptable by requesting more heat or cooling at a suitable user interface. This feedback, or its absence, allows the energy management system to learn a satisfactory operating temperature tolerance which may vary according to the time of day or day of the week in accordance with the behaviour model and the times when user feedback was given.

A similar temperature tolerance range is obtained for hot water. In this case, the maximum temperature of 60 C is determined from the need to sterilise the water held in the hot water tank This temperature must be reached at some time in the day but it is undesirable for the water to be hotter than 60 C for safety reasons. The minimum hot water setpoint for normal use defaults to a suitable value, for example 50 C, and may be varied by user input.

Prediction of energy demands The energy management system of the present invention typically predicts the demand for electricity, hot water, and space heating or cooling for the next 24 hours by performing a computation at a suitable time overnight when demands are low and static. The variable factors that are inputs to this computation are forecasts of the weather and occupant behaviour. The weather prediction may be as simple as an external temperature profile for the next 24 hours based on the readings of the previous day and the time of year, or it may be a comprehensive forecast including wind speed and weather forecast data derived from a radio broadcast or from additional sensor readings such as barometric pressure. I'he energy output data from the previous day from any renewable energy device under supervision, such as a solar water heater or micro wind turbine, for example, may also be taken into account when deriving the weather prediction.

The forecast of occupant behaviour over the next 24 hours is performed using the model generated using the methods described above. In most cases, the dominant driving factor in the model will be the day of the week. A profile of electrical load, -17-hot water demand times and the time settings for space heating or cooling can then be predicted using the forecast of occupant behaviour.

By applying these two forecasts (i.e. weather and occupant behaviour) to the building and hot water storage models, and taking account of the occupant temperature tolerances and capacity for energy storage that have been determined using the methods previously described, the maximum and minimum needs for energy inputs can be computed for each time point in the 24 hour prediction.

Preparation and execution of an optimum schedule Having identified energy needs for the next 24 hours, the energy management system according to the present invention will preferably assign the available energy conversion devices to meet them. Like demand prediction, this is typically performed as an overnight computation. There are many known techniques for optimising the assignment of resources to tasks but all depend on being able to assign a cost to the use of each resource. Because of the wide range of energy sources and energy conversion devices that may be managed by the present system, and its automatic operation (which may mean that it does not have information on the monetary cost per kilowatt hour of energy sources such as gas and electricity), then by default the cost it seeks to minimise is energy loss, calculated in accordance with both first and second laws of thermodynamics. Where information on the monetary cost per kilowatt hour of a particular energy source is available then that information is used to adjust the cost assigned to use of the relevant resources in the optimisation calculation.

In order to calculate the energy loss occurring in the operation of each energy conversion device under supervision, the energy management system holds default values for key parameters such as the calorific value of mains gas and then uses the data collected under the methods previously described to compute the efficiency of the energy conversion device.

In order to assign the correct cost to the use of mains electricity, account must be taken of the typical losses that occur in its generation and distribution. These vary -18-during the day with the highest losses occurring at the time of peak demand in the early evening because the least efficient plant is used to meet the demand peak. In its simplest form the energy management system holds a set of typical loss profiles for mains electricity over 24 hours for the time of year. In a more complex embodiment, suitable for applications where a relatively high consumption of mains electricity is expected, it should preferably be able to receive a radio broadcast such as that in accordance with BS 7647:1993 which can provide up to date information on the cost of electricity. Where electricity is generated locally for example by a CHP generator or solar photovoltaic generator it is considered to substitute for mains electricity to the extent that sufficient demand is predicted, and the additional losses which occur when any surplus is exported to the local mains distribution are taken into account.

In some applications the energy management system may have some electrical appliances under management whose time of operation can be controlled. For example, the building may have a tumble drier which can be scheduled to run at a time when electricity is available at low cost. The energy management system will record the demand level of such electrical appliances on first use so that subsequently their on/off operation can be planned as part of the overall optimised schedule.

Once a cost function is available for the use of each energy source and energy conversion device then an optimisation calculation can be performed using Simplex linear programming or any other known mathematical optimisation technique. This finds the minimum cost schedule by making use of these cost functions in conjunction with the flexibility in operating time offered by the thermal capacity of the building and hot water storage, the storage capacity of any energy conversion device with this capability, and any controllable electrical demands.

Having computed an optimum schedule the energy management system puts it into effect for the planned period. The energy management system continues to monitor sensor measurements and if they indicate a result that is different to the predicted result then the system takes appropriate action. For example, if external temperatures are lower than predicted, or the occupant or occupants of the building present an -19-unexpected demand for hot water, then the additional heat demand is met using the energy conversion device that is able to meet the real-time demand most efficiently.

Employment of default assumptions and simple control methods The knowledge base which this energy management system obtains using the methods described for energy conversion device recognition and modelling, occupant behaviour modelling and building thermal modelling may take several months to reach maturity because all the methods require collection of sensor data over a period of time. It is therefore important that the energy management system is able to operate effectively while this data collection and processing takes place. This is achieved by making initial assumptions based on the lifestyle and comfort preference that is input to the system on installation in a similar way to that described in GB 0522544.6. The lifestyle and comfort preferences can be entered by the user or the installation technician. These lifestyle and comfort preferences tell the energy management system, for example, that the occupant of the building is a single older person normally at home all day who requires a high level of comfort so that the initial time settings for space heating and setpoints will be automatically configured accordingly. Alternatively the lifestyle setting might be for an office whose occupants follow a conventional office hours routine In this case, the bias of the energy management system decision making should be towards economy of operation so time settings and setpoints will be automatically configured economically to suit this particular scenario.

With time settings and setpoints for space heating and hot water heating obtained in this way, the energy management system can operate immediately on start-up in a conventional mode of switching the primary energy conversion devices under direct control on and off to achieve the desired setpoints. Approximate values for the building thermal parameters will be obtained within a few days, so from then on more optimised control decisions can be taken. As occupant behaviour patterns and energy conversion devices not under direct control are recognised then the optimisation of control will improve.

-20 -In order to establish the characteristics of the secondary energy conversion devices, the energy management system will preferably employ them as soon as reasonably possible after installation on an experimental basis to determine their efficiency and other characteristics and thereafter employ them when justified as part of the minimum cost schedule or to meet an unexpected energy need when it arises.

Drawings Figure 1 is a schematic diagram showing how an energy management system can be used to control the heating of a hot water tank by an immersion heater and a gas boiler; Figure 2 is a schematic diagram of an electronic energy management unit that forms part of the energy management system of Figure 1; Figure 3 is a schematic diagram showing how the energy management system can be used to control the heating of a hot water tank by an immersion heater and a combined heat and power (CHP) generator; Figure 4 is a schematic diagram showing how the energy management system can be used to control the heating of a hot water tank by an immersion heater, a combined heat and power (CHP) generator and a solar water heater, and the operation of an air conditioning unit; Figure 5 is a schematic diagram showing how the energy management system can be used to control the heating of a hot water tank by an immersion heater, space heating using both electrically heated thermal storage heaters and radiators, and the operation of a washing machine; and Figure 6 is a schematic diagram of an electronic energy management unit that forms part of the energy management system of Figure 5.

An energy management system according to the present invention will now be explained with reference to Figures 1 to 6.

In a first example shown in Figures 1 and 2, an electronic energy management unit 10 is used to control space heating and hot water heating. Cold water is supplied to the bottom of a hot water tank 2 from a mains water supply as shown. A heating coil 4 is -21 -located in the hot water tank 2. Hot water from a gas boiler 5 can be circulated through the heating coil 4 by a pump 6 to heat the water in the hot water tank 2. Hot water from the gas boiler 5 may also be supplied to radiators (not shown) to provide space heating. The direction of circulation, either through the radiators (not shown) or the heating coil 4, is determined by the valve 7 under the direction of an electronic control unit 26 for the gas boiler. The control unit 26 is used to control the local operation of the gas boiler 5, the pump 6 and the valve 7 and receives primary commands from the energy management unit 10 via an electrical cable 24.

The hot water tank 2 is also fitted with an electric immersion heater 8 that receives secondary commands from the energy management Unit 10 via an electrical cable 22.

In other words, the gas boiler 5 (via the heating coil 4) is configured to be the primary hot water heater and the immersion heater 8 is configured to the secondary hot water heater. The gas boiler 5 is also configured to provide primary space heating. In the arrangement of Figure 1 there is no energy conversion device for providing secondary space heating.

The energy management unit 10 has a first manual switch 12 that allows the user to select a lifestyle from a list of options. A second manual switch 14 allows the user to indicate when the room temperature setpoint is to be raised or lowered.

The energy management unit 10 receives water temperature data in the form of continuous temperature measurements from a pair of sensors 16 and 18 located at upper and lower regions of the hot water tank 2, respectively. It also receives external air temperature data in the form of continuous temperature measurements from a sensor 13 located outside the building and internal air temperature data in the form of continuous temperature measurements from a sensor 15 located inside the building.

An electrical load sensor 17 located between the main building circuit breaker or fuse panel and the electricity meter 20 measures the aggregate electrical load presented by the various electrical appliances within the building and provides electrical load data to the energy management unit 10.

-22 -A gas sensor is integrated with the gas meter 19 and measures the rate at which gas from the mains gas supply is consumed. Similarly, a power sensor is integrated with the electricity meter 20 and measures the amount of power exported from, and imported to, the mains electricity supply.

On start up, the energy management unit 10 adopts initial room temperature and hot water setpoints and time settings for space heating and hot water heating by using the lifestyle setting and the external air temperature measurements provided by the sensor 13 to select from a database of stored defaults. The energy management unit 10 then issues primary commands to the control unit 26 which controls the gas boiler 5, pump 6 and valve 7 to achieve the setpoints for room temperature and hot water during those times when space heating and hot water heating is to be provided. The energy management unit 10 determines the response times of the gas boiler 5 to the primary commands, the gas consumption and the electricity consumption of the gas boiler 5 (typically arising from its internal fan and electronic control circuits etc.) and the pump 6 from the changes in the data provided by the various sensors. After a relatively short period of time (perhaps a few days, for example) the thermal parameters of the building can also be calculated from the energy inputs and the overnight rate of fall in temperature. The time settings (i.e. the start and stop time) for space heating can then be adjusted automatically to take account of the thermal parameters.

If the initial room temperature setpoint and time settings are not to the liking of the occupants they can provide feedback by requesting more or less heat using the manual switch 14 This will cause the daily time-varying temperature setpoint profile to be adjusted accordingly and their acceptable temperature tolerance can be determined.

The daily time-varying temperature setpoint profile will also be adjusted in response to external temperatures by increasing it slightly when the outside temperature is low and vice versa. This allows the energy management unit 10 to use a single temperature reference point which is normally at the centre of the building. The -23 -extremities of the building will usually be closer to the external temperature, so this adjustment helps to ensure the extremities remain within the acceptable temperature range.

Sensor measurements of electrical load and the hot water temperature will allow the energy management unit 10 to build a model of the behaviour patterns of the occupant or occupants of the building within a few weeks if their habits are fairly regular. The model will cover the times when the occupants are active (i.e. in residence and awake), asleep, and absent. It will also include the times when hot water is demanded and the quantities of hot water that the occupants use. The time settings for space heating or cooling and hot water heating will then he adjusted accordingly.

At some time during the first few weeks of operation the energy management unit 10 will meet a hot water demand by issuing a secondary command to the immersion heater 8 rather than a primary command to the control unit 26 for the gas boiler 5.

The electrical energy used by the immersion heater 8 will be measured by the electrical load sensor that is integrated with the electricity meter 20. The energy management unit 10 can calculate the relative efficiency of the gas boiler 5 for hot water heating by comparing the amount of electrical energy used by the immersion heater 8 with the amount of gas that is consumed by the gas boiler 5 when executing a primary command. The volume of the hot water tank 2 can also be determined from the rate of rise of the water temperature when the immersion heater 8 is operating.

The efficiency of the gas boiler 5 for space heating will be calculated over time by comparing the heating effect of the electrical load when that is high with that achieved by the gas boiler. These continuous measurements of gas boiler efficiency are used to provide a prompt for maintenance when the efficiency declines below an acceptable level. For example, an alarm lamp indicator 28 can be lit to notify the occupants in the event that the gas boiler is no longer operating within efficiency limits.

-24 -Once the energy management unit 10 has developed a mature set of mathematical models covering occupant behaviour, the thermal parameters of the building and the available energy conversion devices, then it will be able to produce an optimised operating schedule for the next day including time settings that take account of all these factors, and decisions such as to use the immersion heater 8 in preference to the gas boiler 5. The predicted external air temperature profile for the coming day employed in the optimisation will be based on that recorded for the previous day, and the time of year provided by an internal date and time facility.

In a second example shown in Figure 3, the gas boiler 5 has been replaced by a combined heat and power (CHP) generator 30. The CHP generator 30 uses the same mains gas supply as its energy source but when producing heat under the control of the energy management unit 10 and the control unit 26 it also produces electricity from the generator part 32 which is supplied to the building electricity distribution via the cable connection 31. The energy management unit 10 recognises the presence of a CI-IP generator automatically from the difference between the electrical power measurements provided by the electrical load sensor 17 and the sensor integrated with the electricity meter 20 thaI occurs when a command to provide heat is given. The energy management unit 10 therefore builds up a model for the CHP generator 30 that relates heat demand to electricity generation.

With this model of the CHP generator 30, the energy management unit 10 can create an optimised schedule that takes advantage of the thermal capacity of the building and knowledge of the behaviour patterns and energy demands of the occupant or occupants of the building. For example, if the energy management unit 10 predicts that the occupants will be present and consuming most electricity in the evening due to their use of lighting, cooking and entertainment appliances it will schedule the room temperature setpoint to rise slightly during the day to the extent that is within the known temperature tolerance of the occupants. This will cause the CHP generator 30 to deliver most of the heat requirement of the building in the evening and at the same time generate electricity when there is most demand. Without this optimisation more of the electricity output of the CHP generator 30 would be exported back to the mains electricity supply leading to higher energy losses and monetary cost.

In the example shown in Figure 4, two additional energy conversion devices have been installed in the building. These are an air conditioning unit 36 and a solar water heater that includes a solar panel 34 and a pump 35 for circulating water from the hot water tank 2 through the solar panel where it is heated. The air conditioning unit 36 is powered by electricity from the main domestic electricity supply busbar through the circuit breaker or fuse panel and is controlled by the electronic energy management unit 10 in response to primary cooling commands delivered via an electrical cable 37.

When the air conditioning unit 36 is installed, the energy management unit 10 detects that an energy conversion device for primary cooling is available and operates it when necessary according to a room temperature setpoint and the occupant behaviour model. The cooling needs and temperature tolerances of the occupants are determined initially from lifestyle settings, and subsequently evolved through their use of the manual switches 14 in exactly the same way as heating needs. By measuring the electricity consumption of the air conditioning unit 36 the energy management unit 10 determines the energy efficiency of the air conditioning unit and the dependence of this efficiency on the external and internal air temperatures, The energy management unit 10 can then combine this information with its knowledge of the building thermal capacity, the predicted external air temperatures for the coming 24 hours, the temperature tolerances of the occupants, the behaviour patterns of the occupants, and the cost of electricity at different times to schedule the operation of the air conditioning unit 36 at the most efficient time.

When the solar water heater is installed, the energy management unit 10 recognises its presence from the uncontrolled heating of the water in the hot water tank 2 occurring in a way which correlates with a reference profile for solar radiation intensity as disclosed in GB 0702392.2. More particularly, the energy management unit 10 is able to detect the energy output of the solar water heater from a rising difference between the temperature of the water inside the hot water tank 2 as measured by the sensors 16 -26 -and 18. The temperature measurements from the sensors 16 and 18 are stored in a database (see Figure 2) over a period of time as energy data. The energy management unit 10 uses the stored energy data to derive a profile. This can be carried out by dividing the whole of the stored energy data into consecutive 24 hour periods and then averaging these together to form a single derived daily profile. A fixed read-only database (see Figure 2) contains individual reference profiles (i.e. one for each day of the year) for a series of energy sources. Alternatively, a reference profile for a particular energy source may be calculated on demand using a mathematical formula in combination with a smaller number of reference profiles where necessary. In practical terms this would be useful if storage capacity was limited.

A mathematical correlation function is then used to compare the derived profile (i.e. the profile derived from the stored energy data) against the reference profiles. If the correlation function determines that the derived profile matches a reference profile (for example, if it returns a correlation factor that is greater than a selected threshold) then the energy management unit 10 can use this determination and any other known information to work out the type of energy conversion device. For the particular arrangement shown in Figure 4 then the correlation function will match the derived profile with a rctèrence profile for solar radiation intensity and the energy management unit 10 will identify the previously unknown source of water heating as a solar water heater, A similar process of identification and characterisation can be applied to other types of energy conversion device, including those such as gas boiler and CHP generators that use an auxiliary energy source like mains gas. The output of the relevant process block of Figure 2 is therefore a recognition of the presence of a particular type of energy conversion device that is to be taken into account in the decisions of the energy management unit 10.

Over time, the energy management unit 10 builds a model of the water heating that can be expected from the solar water heater under different weather conditions and at times of the year. The energy management unit 10 then includes this predicted output in the planning of an optimised schedule for the controllable energy conversion devices that can provide water heating (i.e. the CHP generator 30 and the immersion heater 8).

In the example shown in Figure 5, a modified electronic control unit 40 has additional sources of information available to it. For example, the electronic control unit 40 includes a radio receiver Ii that allows it to pick up an overnight broadcast of the varying cost of mains electricity at different times during the coming 24 hours. This broadcast may comply exactly with BS 7647:1993 or may by a more sophisticated development of that concept. The energy management unit 40 is also equipped with a barometric pressure sensor 45 which is used with the external air temperature data and the time of year to generate weather forecasts for the coming day, possibly in combination with weather forecast information included in the radio broadcast.

Additional or alternative weather sensors may be optionally provided. Other aspects of the electronic control unit 40 are as described above.

The improved information allows the energy management unit 40 to make more accurate forecasts than was possible with the previous examples. The resulting changes to the information flows with respect to the internal functionality of the energy management unit 40 are shown in Figure 6.

The energy conversion devices under the control of the energy management unit 40 include electrically heated thermal storage radiators 41, electrically heated radiators 44 and an immersion heater 8. A washing machine 46 may also be operated at certain times under the control of the energy management unit 40. Commands for primary room heating, secondary room heating, energy storage and electrical appliance control are transmitted from the energy management unit 40 over electrical cable 42 to local controllers 43 which execute the commands for the device to which they attached. The delivery and execution of these commands may use any of a range of

known technologies such as the LonWorks system, for example.

On first use of the energy management unit 40, if the time of year and external temperatures indicate that space heating will be needed it will issue an energy storage -28 -command to turn on electricity to the storage radiators 41 at the earliest opportunity when the cost of electricity is below a default threshold, and allow them to charge until they turn off themselves due to their capacity being reached. During the charging process, the energy management unit 40 will learn the charge rate, the maximum amount of energy that can be stored and the increase in room temperature caused by heat "leakage" from the thermal stores. When heating is required then the energy management unit 40 will control the storage radiators 41 as the primary source of space heating and determine their characteristics accordingly. From then on the overnight calculation of an operating schedule will include an accurate determination of how much energy needs to be stored for the coming day given the weather forecast, the thermal properties of the building and the behaviour patterns of the occupant or occupants of the building. The required energy will be taken as electrical power at the lowest cost time as determined using the information in the radio broadcast. This precise determination of the energy charge will provide significant energy savings and increased comfort for the user compared to the conventional method of charge regulation based on room temperature disclosed in GB 2265454.

Under colder weather conditions the storage radiators 41 may have insufficient stored energy to maintain the required room temperature in the evening In this case the secondary radiators 44 are brought into use by the energy management unit 40. As disclosed in GB 2384300, the use of secondary heating devices which emit a greater proportion of radiant heat gives a perception of greater comfort than the same amount of energy provided simply as convection. If the occupants are able to tolerate a lower room temperature when the secondary radiators 44 are in use this will be detected by the energy management unit 40 and the resulting gain in energy efficiency will be taken into account in calculating an optimum operating schedule.

The immersion heater 8 is controlled using the primary hot water heating commands via electric cable 22. Its operating time is scheduled to minimise cost taking account of the occupants' usage patterns, the cost of electricity as indicated in the radio broadcast and the insulation properties and capacity of the hot water tank.

Claims

-29 - CLAIMS

1. An energy management system for efficient and automatic management of one or more energy conversion devices located in a building, the system comprising: a plurality of sensors measuring energy related parameters such as temperature and electrical load and providing energy sensor data, device recognition means for automatic recognition, characterisation and modelling of the one or more energy conversion devices from the energy sensor data; thermal modelling means for automatic characterisation and modelling of the thermal properties of the building from the energy sensor data; occupant behaviour recognition means for automatic recognition of whether or not the building is occupied at any particular instant in time and for generating a predictive model of' the behaviour patterns of' the occupant or occupants of the building; temperature tolerance identification means for identifying the temperature tolerances of the occupant or occupants of the building from their response or absence of' response to temperature setpoints that are varied automatically; energy demand prediction means for predicting energy demands depending on the weather, building occupancy and behaviour patterns of the occupant or occupants of the building; and energy management means to use the information obtained by the device recognition means, the thermal modelling means, the occupant behaviour recognition means, the temperature tolerance identification means and the energy demand prediction means to schedule the operation of the one or more energy conversion devices in an efficient manner thereby meeting the predicted energy demands.

2. An energy management system according to claim I, wherein the energy management means uses initial values and control models that are progressively improved as more energy sensor data is accumulated over time so that energy -30 -demands can be predicted more accurately and the operation of the one or more energy conversion devices can be scheduled in a more efficient manner.

3. An energy management system according to claim 2, wherein the initial values and/or the control models are derived from user data.

4. An energy management system according to any preceding claim, wherein the energy management means schedules energy demands that can he shifted in time so that they take place when they can make the most efficient use of the one or more energy conversion devices and any available energy sources.

5. An energy management system according to any preceding claim, wherein the device recognition means uses the energy sensor data to derive a profile of the changes over time in the energy related parameter, correlates the derived profile with one or more reference profiles, each reference profile corresponding to a particular energy source, and uses the results of the correlation to identify the type of energy conversion device.

6. An energy management system according to any preceding claim, further comprising means for predicting the energy output of renewable energy sources depending on the weather, time of year or time of day.

7. An energy management system according to any preceding claim, wherein the energy management means uses information about the future cost of electricity at different times of day.

8. An energy management system according to claim 6, wherein the information about the future cost of electricity is provided by a broadcast signal. -31 -

9. An energy management system according to any preceding claim, wherein the energy management means is able to identify a deterioration in performance or need for maintenance of any of the energy conversion devices.

10. An energy management system according to any preceding, wherein the temperature setpoints used by the temperature tolerance identification means are time-varying temperature setpoints.

11. An energy management system according to any preceding claim, wherein the temperature setpoints used by the temperature tolerance identification means are a hot water temperature setpoint and a space heating temperature setpoint.

12. An energy management system according to any preceding claim, wherein the temperature setpoints are adjusted with reference to the external air temperature.

13. An energy management system according to any preceding claim, where the operation of an energy conversion device is controlled by the energy management means in response to the detection of an abnormal thermal load.

14. A method of energy management for efficient and automatic management of one or more energy conversion devices located in a building, the method comprising the steps of: measuring at least one energy related parameter to obtain energy sensor data; automatically recognising, characterising and modelling the one or more energy conversion devices from the energy sensor data; automatically characterising and modelling the thermal properties of the building from the energy sensor data; -32 -automatically recognising whether or not the building is occupied at any particular instant in time and generating a predictive model of the behaviour patterns of the occupant or occupants of the building; identifying the temperature tolerances of the occupant or occupants of the building from their response or absence of response to temperature setpoints that are varied automatically; predicting energy demands depending on the weather, building occupancy and behaviour patterns of the occupant or occupants of the building; and scheduling the operation of the one or more energy conversion devices in an efficient manner to fleet the predicted energy demands.

15. An energy management system substantially as herein described and with reference to the drawings.