BACKGROUND OF INVENTION

The present application relates to a water heater controller.

Electrical water heaters are commonly used to heat water for household consumption in developing countries where gas is not readily available.

South Africa is one such country, and boasts 5.4 million electrical water heaters. Similar to many developing countries, South Africa's national electricity utility is unable to meet the energy demands of the country and must cut service provision in certain areas through load shedding during periods of high demand to ensure that the generation capacity of the grid is not exceeded.

Water heating is responsible for 7% of the country's demand, and 20% of the residential demand. However, during peak hours, it constitutes between 30% and 50%.

Part of the energy consumed by water heaters is to replenish heat dissipated to the environment.

This type of energy is referred to as standing losses, and could be as much as 20% of the water heater's consumption. These standing losses can be virtually eliminated if a timer control is applied to only heat the water before warm water is needed.

Demand side management (DSM) aims to flatten utilities' demand curve (e.g. peak shaving and valley filling) by shifting customer energy usage and reducing losses on the load side. This is advantageous to utilities as it allows for the deferral of infrastructure development to increase generation capacity by instead reducing the demand.

Water heaters are well-suited to DSM programs as they are able to store energy. However, many of these devices are mismanaged and suffer from large standing losses as warm water is available throughout the day, even for extended periods where no usage occurs.

DSM control techniques and programs have been created to more effectively manage the energy consumption of residential water heaters.

However, for these controllers or programs to be effective, an accurate water usage profile is essential to coordinate the switching times for water heaters. This is because consumer usage patterns vary between users, seasonally, and between regions.

For example, in South Africa, it was found that warm water consumption increased by up to 70% from summer to winter and that high-income households consumed up to four times more warm water than low-income households.

If generic assumptions are made about these patterns of use, they may be inaccurate and result in consumption being adversely affected.

For indirect load management programs, where consumers are responsible for the control of their devices, customer participation is important. Users need to be able to control and understand their energy consumption in a simple and convenient manner. This is not currently the case with water heaters which are usually positioned in hard to reach locations (such as on roofs or in attics). Additionally, users don't always know the best means of controlling their water heaters for energy savings.

For example, they may not know when to switch the water heater on and off to reduce energy consumption but still have hot water on demand when needed.

The present invention seeks to address this. SUMMARY OF INVENTION

According to one example embodiment, a water heater controller includes: a memory for storing data therein, the data including historical water usage data including the time, date and amount of historical water usage from at least one water heater to be controlled by the controller, the data further including water heater energy loss data; and a processor connected to the memory, the processor: retrieving the historical water usage data from the memory for a past period of time and using this to determine a future water usage schedule; and retrieving the water heater energy loss data from the memory and using this together with the future water usage schedule to calculate a future energy supply schedule for the at least one water heater so that hot water will be available when required as determined by the future water heater schedule whilst the amount of energy required to provide the hot water will be reduced. According to another example embodiment, a method of controlling a water heater includes: storing data in a memory, the data including historical water usage data including the time, date and amount of historical water usage from at least one water heater to be controlled, the data further including water heater energy loss data; and retrieving the historical water usage data from the memory for a past period of time and using this to determine a future water usage schedule; and retrieving the water heater energy loss data from the memory and using this together with the future water usage schedule to calculate a future energy supply schedule for the at least one water heater so that hot water will be available when required as determined by the future water heater schedule whilst the amount of energy required to provide the hot water will be reduced.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic view of a water heater which will be controlled according to an example embodiment;

Figure 2 shows a schematic view of an electronic controller used to control the water heater illustrated in Figure 1 ;

Figure 3 shows a schematic representation of a mobile telephone for use with the present invention;

Figures 4 show displays of information on the screen of the mobile and 5 telephone. DESCRIPTION OF EMBODIMENTS

The present invention relates to a water heater controller that uses historical water usage for a past period of time to determine a future water usage schedule. It then uses a complex water heater energy model which will be described below to determine an efficient future energy supply schedule for at least one water heater so that hot water will be available when required, as determined by the future schedule, whilst the amount of energy required to provide the hot water will be reduced.

By way of background, and referring to Figure 1 , a standard electrical water heater 10 is shown which includes a cold water inlet 12 and a hot water outlet 14.

These water heaters are well-known. Water flows into the inlet 12 and is heated in the body of the water heater 10. On user demand, the water flows out of the water outlet 14 through pipes to a tap where the hot water is used, typically for a bath or shower or possibly to wash dishes or clothes.

The water heater includes an element 18 which is controlled to heat water contained inside the water heater 10.

In order to implement the present invention, the historical water usage needs to be captured which will then be used to determine a future required water usage schedule.

This could be done in a number of ways, one of which is to place a water flow meter 16 on the inlet pipe as when water is removed from the water heater, cold water flows into the water heater via the inlet pipe.

Another methodology is to use a temperature sensor 20 which is placed on the hot water outlet 14 to measure the temperature of the outlet, where temperature fluctuations sensed by the temperature sensor 20 are used to determine start and end times of water usage events.

Whether a water flow meter 16 or a temperature sensor 20 are used, the objective is to capture historical water usage events including the time and date of the water usage and the amount of water used as this will be required to determine the future water usage requirements from the water heater 10.

Referring to Figure 2, electronic water heater controller unit 22 includes a processor 24 and a memory 26.

The electronic controller unit 22 will typically be located in proximity to and connected to the water heater and will control a single water heater.

However it will be appreciated that in a more distributed system, an electronic controller unit 22 could be located remote from the water heater it is controlling. In a further embodiment, a single electronic control unit 22 could be set up to control a plurality of water heaters.

In any event, the electronic controller unit 22 also includes a communications module 28 which serves a variety of functions that will be described below.

Once detected, using either of the methods referred to above, historical water usage events for an amount of time are stored in the memory 26 and these can now be used to determine a more energy efficient control cycle for the water heater 10 whereby water is only heated up when the processor 24 determines that hot water will be required by the user of this particular water heater.

The processor 24 retrieves water usage events, from the memory 26, for a past period of time. These events are then used to determine a water energy supply schedule for future water usage requirements from the water heater.

Water heater energy loss data is also retrieved from the memory 26 and this is used together with the future water usage schedule to calculate a future energy supply schedule for the water heater so that hot water will be available when required as determined by the future schedule whilst the amount of energy required to provide the hot water will be reduced.

In a further aspect of the present invention, the processor 24 interacts with the user of the water heater 10 via communications module 28.

Data is transmitted and received via communications module 28 to a communications device 30 of the user.

It will be appreciated that the communications device 30 of the user could be any suitable communications device such as a mobile telephone, a tablet or a laptop or desktop computer.

A prototype of the present invention was implemented using an executable application on a mobile telephone of the user.

Referring to Figure 3, a mobile telephone 30 is illustrated which includes a communications module 32 and a processor 34. The mobile telephone also includes a memory 36.

In addition, the mobile telephone includes a display 38 by means of which information is displayed to a user, and the user interface 40 by means of which user inputs can be obtained.

In many mobile telephones the display 38 and user interface 40 are implemented by way of a touchscreen with which the user interacts. In any event, an executable application is downloaded onto the mobile telephone 30 and executes on the processor 34.

In the illustrated example embodiment, the mobile telephone 30 will receive from the electronic controller unit 20 data including the user's historical water usage data as calculated by the processor on the electronic controller unit 22 including a proposed future schedule.

The data is displayed to the user via the display 38 and the user is prompted for input regarding the future schedule.

Thus the user may alter the future schedule and the alterations will be transmitted from the mobile telephone 30 back to the electronic controller unit 22.

In addition, the user is able to input to the executable application to override the schedule at any time and switch the water heater on or off.

In addition, the user is able to provide input to the executable application in order to alter the temperature to which water will be heated in the water heater.

Figure 4 shows an example optimise display that is displayed to the user via the display 38.

In this example embodiment, in order to generate a recommended schedule, the user is prompted to select a day that represents their typical usage. The user selects a date and the water usage events for that date are retrieved from the electronic controller unit 22 and displayed to the user as illustrated in the Figure.

For accurate data capture purposes, a list of detected water usage events are displayed (Figure 4), which will allow the user to select to include or exclude the event in the optimisation calculation. For example, the user will know if the event was a shower or a bath, or a spurious unintended use of warm water. This data input by the user will be transmitted back to the electronic controller unit 22 for more accurate data capture.

The electronic controller unit 22 could use this information by allocating a specific amount of energy that is typically required by a certain type of usage event. For example, a shower requires a warm water flow rate of 7 litres per minute for a warm water temperature of 65 °C.

Depending on the set temperature of the water heater, the flow rate for a specific event could be scaled up or down to provide the required energy to produce a usage event using the first law of thermodynamics. This flow rate can then be used by the electronic controller unit to determine the amount of energy consumed by a usage event when providing an estimate of the change in energy consumption of the water heater for various schedules and set temperature settings.

In any event, referring back to Figure 4, the user is also able to select the events that are part of their daily routine, from the list of events that are detected by selecting the "Include" boxes on the right of the display.

In the illustrated example, the user has selected three events out of the seven events displayed thereby instructing the electronic controller unit 22 to only use the three events to calculate a future schedule.

Thus the communications module 28 receives back from the user's communications device 30 a selection of which of the detected water usage events to include in a revised future water usage schedule and wherein the processor 24 will use the revised water usage schedule together with the water heater energy loss data to calculate a revised future energy supply schedule for the water heater. Data is transmitted to the electronic controller unit 22 which calculates a proposed future schedule and transmits this back to the mobile telephone 30 where it is displayed to the user at the bottom of the display of Figure 4.

In addition to the above, the energy consumption data and likely energy- saving data can also be displayed to the user. It will be appreciated that the present invention thus provides the user with the ability to see how much they will save/spend by a change in their geyser's control schedule, and/or by a change in their geyser's set temperature, both for a nominal or measured usage pattern.

Figure 5 illustrates a control display whereby the user can manually implement further changes to the schedule or set the temperature of the water heater via the control tab.

Thus the future energy supply schedule is transmitted by the communications module 28 to the user's communications device 30 to be displayed to the user via display 38.

At the bottom of the display is shown the on/off schedule for the water heater allowing the user to delete or add active slots to heat water or not.

In this way, the communications module 28 receives back from the user's communications device 30 a user input via user interface 40 to alter the future energy supply schedule by increasing or decreasing the time that the water heater is switched on.

The controller of the water heater determines when to switch on and off the water heater to achieve the desired temperatures at the desired times based on the following algorithms.

All the water in the electric water heater (EWH) tank is treated as a single body. This implies that the temperature of the water inside the tank is uniform. Therefore, when a usage event occurs, the water leaving the tank through the outlet pipe is at the average temperature of the water inside the tank (T _inside ). The cold water entering the tank from the inlet pipe, to replace the water used, instantaneously mixes with the water inside the tank to create a new average temperature.

Usage losses refer to the energy lost as a result of hot water usage events (e.g. showering or filling sink for washing dishes). This energy must be replaced by the EWH element to reheat the water inside the EWH to the set temperature. The water leaving the tank through the outlet is at the temperature of the water in the tank and the cold water entering the tank is at the inlet temperature (T _inlet ).

Furthermore, the water entering the tank is assumed to instantaneously mix with the remaining hot water in the tank to create a new average internal temperature. Under the assumption that water at 20 °C is the baseline for zero energy, all the energy inside the EWH is then held by the remaining hot water inside the tank and the resultant temperature of the water in the EWH at time t is given by the energy balance equation:

^inside ⁼ ^hot

Where: E _inside is the energy inside the EWH tank; and E _hot is the energy in the remaining hot water inside the EWH tank. This results in the following equation.

cpVtank[Tinside(t) ~ ^T inlet] = ^c P ^V hotIThot ~ Tinlet]

Where: c is the specific heat capacity of water (4180 J · (kg · K) ^_1 ); p is the

kg

density of water (1000 ^); T _inside (t) is the average temperature of the water inside the EWH at time t; V _tank is the total volume of water in the EWH; and V _hot and T _hot are the volume and temperature of the unused hot water remaining inside the EWH respectively. Cancelling out the constants and solving f or T _inside (t):

^inside ^{= h0} ' (Tj, _ot - T _in i _et ) + Tj _nlet ... 1 The energy (heat) lost during a usage event (E _usage ) can then be obtained as follows:

Eusage = cm_yr

· ^" ' E _US age ⁼ ~ Rafter) ·· · 2

Where: V is the volume of water used (equal to the volume to be heated); and ΔΤ is the temperature change that is required to reheat the water in the EWH from the average temperature after the usage event (T _after ) to the temperature it was before the usage event (T _{be ore} ).

From equation 2 it can be seen that the energy required to raise the temperature of the water in the tank is dependent on the change in temperature required, and not the absolute temperature of the water. This implies that the same amount of energy is required to increase the temperature of the water from 50 to 51 °C as is required to raise the temperature from 60 to 61 °C.

Standing losses refer to the energy lost due to heat dissipation from the water inside the EWH to the outside environment as a result of the temperature difference between them.

The maximum allowable standing losses over a 24 hour period, as stipulated by South African National Standard (SANS) 151 , for a closed type 150 litre EWH is 2.59 kWh at a set temperature of 65 °C [SANS 151].

The worst case value of the thermal resistance (R) of the EWH tank can then be calculated as follows :

_ 1

Qstanding ~ ^ ( inside ~ ^ambient)

Solving for R and substituting the maximum allowable standing loss: 1

R (Tinside - T _ambie „t) = (65 - 20) = 17.4 0 ^■ (kWh) ^"1 · day

Qstanding

This results in a water heater thermal conductance of:

This value concurs with those used in literature. The temperature decay of the water inside the EWH tank toward the ambient temperature is described by: cmTjnsicjett) - [Tjnside 00 T _aml) j _ent ]

Solving this equation for T _inside (t) : T _amlj j _ent ]e

Equation 3 illustrates how the internal temperature of the EWH decays exponentially from its initial value at t = 0 towards the ambient temperature over a given timet. The energy lost to the environment can then be obtained using:

^standing ~ ^cm tank^

" '- ^■ standing ~ Cp tank [Tinside (0) ~ ' TinsideCt)] -

Therefore, the standing losses are given by the amount of energy needed to reheat all the water in the EWH tank to its initial temperature at t = 0.

The goal of the EWH system is to maintain the temperature of the water inside its tank at a pre-set value. The temperature of the water inside the tank is monitored by a thermostat inside the tank. The thermostat is modelled as a switch with hysteresis. If the thermostat observes that the temperature in the tank is lower than the lower set point temperature (Tiower). it will switch on the EWH element. Once the thermostat detects that the upper set point temperature (T _upper ) has been reached, it will switch the element off. The values of the upper and lower thermostat were set to 1 °C above and below the set temperature of the EWH, respectively.

Heat is produced when the element of the EWH is switched on, which will increase the temperature of the water inside the EWH. For the model currently being described, it is assumed that the energy input by the element is distributed uniformly to all the water in the EWH tank. The temperature increase in the water in the tank as a result of energy input by the element (E _input ) over a time interval t can be calculated using:

Ejnput = cmAT

Solving for ΔΤ:

ΔΤ = -^EHL ... 5

cpVtank

This value is then added to the average temperature of the water in the tank.

The EWH system of the present invention allows the user to specify a schedule that only allows the element to be switched on during certain time intervals. A day is broken down into 15-minute intervals for which the EWH is set to be either: active, where element is switched on and off to maintain the set temperature; or inactive, where the element remains off regardless of the present temperature of the water inside the tank.

The purpose of using a schedule for the EWH system is to minimise the energy lost due to standing losses by reducing the temperature of the water in the geyser when no hot water is required. The water can then be heated before usage events to still give users access to hot water on demand.

Although energy is used to reheat the water that cools down when the EWH is inactive, there can still be a net decrease in the energy usage of the EWH through a significant reduction in the standing losses of the EWH. The energy input and outputs are shown in the table below for the different states of the EWH.

Additionally, the smart EWH system also allows the set temperature of the EWH to be selected by the user. Decreasing the set temperature of the EWH results in further reduced standing losses and, therefore, energy savings for users while maintaining the temperature of the hot water at a comfortable level. For example, the EWH is still able to supply a shower with water between 37 to 40 °C for a user to have a warm shower.

If the upper set temperature is exceeded, then the model corrects this overshoot by adjusting the ratio of time that the element was on for during a time interval.

Since the energy input into the system over time, and hence the change in temperature over time, is linear, we can determine the ratio of the time interval (t _rati0 ) that the element was switch on for as follows: Tset(upper) Ti _{n t} j _a i

Tfinal - ^initial

Where T _initial and T _final are the water temperature in the tank initially and at the end of the time interval, respectively.

The value of the energy input by the element over a time interval t is then adjusted as follows:

^input(partial) ⁼ ^ratio ^x Qinput

= tratio ^x (tfinal ~ to) ^x Qinput

Where: Q _input is the power rating of the EWH element; and t _fjnal and t ₀ are the final and initial times of the time interval under consideration.

A two node model is required where some water has been taken out of the water heater and so cold water has moved into the water heater meaning there is a body of hot water and a body of cold water contained inside the heater until the cold water heats up.

For the two node model, the one node model is implemented until the energy in the tank drops by a significant volume in a limited time. After this usage threshold has been exceeded, the water in the tank is divided into two separate nodes. The upper node consists of the hot water that is left in the tank while the lower node consists of cold water from the inlet - mimicking the natural stratification that occurs in the EWC.

The model will transition from the one node to the two node state after 30 litres of hot water has been used in a six minute interval. Any usage event that occurs during this state leaves the outlet pipe at the temperature of the upper node (T _upper ). Water will continue to be supplied at this temperature until the volume of the upper node (V _upper ) reaches zero (i.e. all hot water in the tank has been depleted). Additionally, when a usage event occurs, the volume of water leaving the upper node (V _usage ) is replaced by water entering the lower node through the inlet pipe. It is assumed that the water entering the tank through the inlet pipe mixes instantaneously with the lower node. The temperature of the lower node (T _lower ), after mixing with the water entering the tank at the inlet temperature (T _inlet ), can be obtained using:

_T ■.·<. _ Vlower( ^t )~ ^v usage

ower - 7, 7 L Mower W ~ MnletJ ⁺ Mnlet · ^■ · ¹

v lower W

If the upper node is depleted and the lower node's temperature is higher than the temperature of the water from the inlet then the lower node will become the upper node of the geyser and the new lower node will consist only of cold water at the inlet temperature.

In the two node state, we consider the standing losses and subsequent thermal decay of the two nodes separately. Energy transfers are included in the two node state, including: the standing losses for the upper node ( ^E ioss(upper)) and lower node (E _loss(lower )); the energy transferred from the upper layer to the lower layer as a result of their temperature difference (Econduction); and the energy input by the EWH element (E _input ) .

The thermal resistance used to calculate the standing losses and thermal decay of a node is dependent on the surface area of the node exposed to the environment. This surface area for a horizontal EWH consists of the area of the circular segments on either side of the cylinder, which are identical; and the area of the rectangle that makes up the portion of tank wall for a particular node. In order to calculate the surface area of this rectangle(A _rectangle ), we will need the arc length (s) of the circular segment of each node.

The length of this arc is given by the following:

s = r0 Where: r is the length of the radius of the cylinder; and Θ is the central angle (in radians). The area of the circular segment (A _segment ) can be calculated as the difference between the area of the sector (A _sector ) and the isosceles triangle (A _isosce i _es ) .

Therefore, the area of the circular segment is given by:

•^segment ⁼ A _sect:or - Aj _sosce | _es

= ir ² (0 - sin6) = ir ² [f - _S mg)] ...2

Equation 2 has no analytical solution but a numerical solution can be found using the Newton-Raphson method. Whenever a usage event occurs, we know the volume of water in the tank that has been consumed. From this volume we can calculate the surface area of the circular segment of the lower node using:

V "usage - ~~ I x " Δsegment

Where L is the length of the EWH tank. Then:

For the Newton-Raphson method we define the function f(x): f(x) = C -→ 3Ϊη ζ) = 0

We then set x = ^ to obtain the following:

f(x) = C - x + sin x

f'(x) = - 1 + cosx

A graph of the surface area of the circular segment versus the arc length of the segment for Θ = 0 to 2π is plotted and from this it is seen that the arc length is a monotonically increasing function over this range. An estimate of the arc length can then be obtained using the following equation with an initial estimate x ₀ = π :

f(x _n )

f'(x„)

Once the solution forx has been determined, the arc length s can be obtained by multiplying by the radius r.The surface area of the surrounding rectangle of the cylinder can then be calculated using:

■^rectangle - L ^x S

The exposed surface area of each node is then given by:

^exposed - 2A _se g _ment + A _rectan gi _e

This surface area can then be used to determine the thermal conductance value of each node as follows:

1 1000

^J node

R 24 · A _eX p _0sec)

Where R is the value of the total thermal resistance of the EWH. The thermal decay of each node into the atmosphere can then be calculated using:

TnodeOO = T _ambient + [T _node (0) - T _ambient ]e cmR _node

Where: T _node (t) and T _node (0) are the respective final and initial temperatures of the node; and R _n0 de ^{is tne} thermal resistance (i.e. the inverse of thermal conductance) of the node. The standing losses for a given node (E _standing ( _node )) can then be obtained using: ~~ ^c P ^V node Pnode(t) ^{~ T} node( ^u )] Where V _node is the volume of water in the node under consideration.

While in the two node state, all energy input by the EWH element (E _input ) is assumed to be transferred to the lower node alone. This will increase the temperature of the water in the lower node as follows:

^input ^{= c} PVlower^T

_ΔΤ _ ^E »nput

Cp lower

This temperature is added to the initial temperature of the bottom node. If the temperature of the lower node equals or exceeds that of the top node, the two nodes are merged and the model returns to the one node state.

In order to model the transition from the two node state back to the one node state (in the absence of heating from the element), the stratification within the tank is decayed over time by transferring energy from the upper node to the lower node. The following equation is used to model the conductive heat transfer:

^conductive ⁼ h ^x A _sur f _ace x ( _U pp _er - Tj _ower ) * At

Where: h is the conductive heat transfer coefficient; and A _surface is the cross-sectional surface area between the upper and lower nodes. In order to calculate this surface area we require the chord length given by:

Xchord = 2R x sin (^j

The cross sectional area between the two nodes is then simply the area of the rectangle given by chord length multiplied by the length of the tank. The one-node and two-node models described above unique due to the following:

1. The model is energy based, rather than temperature based, making it much more accurate as it is independent of a problematic internal tank temperature measurement.

2. The model allows for standing losses even in the two-node state

3. The model takes into account scheduling as other models assume an always-on element that is only temperature controlled.

4. The model does energy estimations to determine savings due to changes in control scheme.

Using the above formulae the water heater can be controlled to switch on and off at appropriate times to achieve the hot water schedule requirements in an energy efficient manner.

This is accomplished as follows, the processor 24: a. Knows the desired outlet temperature of the water in the EWH (important to know because that is what reaches the bath/shower/basin, and must meet user's shower/bath expectations)

b. The energy in (i.e. average temperature of) the water in the EWH at a given time, to allow optimised control of it with regards to energy usage. c. The electrical energy usage based on a set control schedule (pool timer-type control), set temperature (from 50 deg C to 75 deg C), and consumption pattern (used to advise a user on which control schedule to employ).

d. The estimated cost of a single bath/shower/other event.

e. The estimated savings achieved in electrical energy and Rand if a hypothetical control scheme (schedule and temperature) is employed in comparison with the existing scheme, assuming the same usage pattern. One example embodiment works on the assumption that the user has selected their preferred water heater set temperature with the mobile phone application.

The communications module 28 receives from the user's communications device 30 a selected water temperature and the processor 24 uses the selected water temperature to determine the future energy supply schedule as described above.

The control system then uses the model to calculate how long it takes the water heater to reach the set temperature, and then applies power to the element just long enough before the first detected event to ensure that the set temperature is reached before the first event.

If there are multiple events required, the model looks at the time separation between the events, and if the volume of water used by preceding events and separation in time allows for the second event to also have hot water, it does not heat in between. If, however the second event will have cold water if extra heating is not provided, it applies more energy (switches element on for longer) to ensure the temperature is high enough before the second event.

The element of the water heater is either controlled by the hardware thermostat if the control mode is set to "Always On". Alternatively, the element is switched on/off by a software thermostat controlled by the local controller if the control mode is set to "Schedule Control".

The controller monitors the tank temperature using a temperature sensor and continues to supply energy to the element until the desired temperature is reached.

It is also important to note, however, that if the software thermostat attempts to increase the water temperature in the tank above the set temperature of the hardware thermostat, the hardware thermostat will override it and turn the element off, essentially controlling the upper bound of the tank temperature.

It will be appreciated that the mobile phone application is used to control and monitor an electrical water heater.

The mobile phone application allows the monitoring (display to user) of: a. The warm water usage patterns, based on water flow meter readings; and

b. Savings made through changes applied to the set temperature and control schedule. (This uses the mathematical model described above.)

The mobile phone application allows control of:

c. The on/off schedule, including an optimised control schedule based on usage patterns; and

d. The set temperature;

as has been described above.

Thus it will be appreciated that the measuring of the temperature on the outlet pipe allows the event detection algorithm to identify warm water usage events using only the outlet temperature reported by the temperature sensor attached to the outlet pipe of the water heater.

The identified consumption patterns can then be used to create an optimised control schedule for users. This is done by allowing the water heater element to turn on only for a period of time before expected usage events occur, which significantly reduces standing losses.