WO2013132239A2

WO2013132239A2 - Temperature assessment

Info

Publication number: WO2013132239A2
Application number: PCT/GB2013/050529
Authority: WO
Inventors: Peter Roberts; Edwin CARTER; Atul K PATEL
Original assignee: Passivsystems Limited
Priority date: 2012-03-07
Filing date: 2013-03-04
Publication date: 2013-09-12
Also published as: GB201204037D0; GB2500034A; WO2013132239A3

Abstract

The invention provides a method of assessing remotely the temperature on the remote side of a barrier in the form of an insulated wall of a hot water cylinder (10), comprising applying to the near side (14) of the barrier a temperature measuring device (22) comprising a first temperature sensor (24) in contact with the near side of the barrier and a second temperature sensor (26) spaced from the first temperature sensor and separated therefrom by thermally insulating material (28), measuring the temperature (θ₁) at the first temperature sensor and the temperature (θ₂) at the second temperature sensor, and calculating from the measured temperature values the temperature (θU) on the remote side of the barrier, based on the thermal conductivity properties of the barrier (U_B) and of the thermally insulating material of the temperature measuring device (U_M). The invention also provides a hot water cylinder, a temperature measuring device for use with a hot water cylinder, and a kit of parts for retrofitting to a hot water cylinder.

Description

Title: Temperature Assessment Field of the invention

This invention concerns temperature assessment, with particular application to hot water cylinders.

Background to the invention

In advanced hot water control systems, temperature sensors are placed on the surface of hot water cylinders (tanks) to provide real-time feedback of the temperature of hot water at strategic heights. These temperatures are then used by the system to determine how much hot water the cylinder contains, eg as a percentage of volume or a relative energy level indication, and also to provide this information to the home occupants as a real-time hot water level indicator.

Hot water cylinders are available in different forms, with varying levels of insulation. The simplest is a copper cylinder with no insulation, in which case temperature sensors can be mounted directly to the cylinder surface to measure the hot water temperature at strategic heights. Hot water cylinders are now more commonly available with an insulated wall including foam or another form of insulation; thus in order to measure the surface temperature of the cylinder, the insulation layer must be removed to create pockets for mounting temperature sensors directly onto the cylinder surface. This is problematic for several reasons:

1) Cutting holes in the cylinder insulation is very destructive and not easy to reverse.

2) Home owners have expressed concern and indeed in rented accommodation this may require landlord's approval.

3) Thermal insulation properties of the cylinder insulation will be slightly degraded.

4) Installation time can be significant, therefore increasing overall costs.

5) Self installation by a home owner is not possible due to the complexity of the process.

6) It is difficult to achieve a thermal bond that will continue to provide accurate temperature readings for many years.

Increasingly, cylinders are constructed to include an outer metal skin which sandwiches an insulating foam layer with the surface of the hot water cylinder; these are commonly known as double-skinned cylinders. This construction makes mounting sensors on the inner cylinder surface skin extremely difficult, as it causes significant damage to the outer skin and invalidates warranties.

This invention addresses the problem of measuring the temperature of a hot water cylinder without removing or destroying the insulation layer.

Summary of the invention

In one aspect, the invention provides a method of assessing the temperature on the remote side of a barrier in the form of an insulated wall of a hot water cylinder, comprising applying to the near side of the barrier (i.e. the outer surface of the insulated wall) a temperature measuring device comprising a first temperature sensor in contact with the near side of the barrier and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material, measuring the temperature (θι) at the first temperature sensor and the temperature (θ₂) at the second temperature sensor, and calculating from the measured temperature values the temperature (9_U) on the remote side of the barrier (i.e. the inner surface of the insulated wall), based on the thermal conductivity properties of the barrier (U_B) and of the thermally insulating material of the temperature measuring device (U_M).

Given a knowledge of the thermal conductivity properties of the barrier and the thermally insulating material of the temperature measuring device (absolute values or the relative values of the two structures, in the direction of heat flow) an estimation of the temperature on the remote side of the barrier (and thus the temperature of adjacent water in the tank) can be calculated as follows. The calculation assumes that all heat flowing through the barrier also flows through the thermally insulating material of the temperature measuring device, i.e. that the two rates of heat flow (6_H) are the same, and that the system is in a steady state with no temperature fluctuations. According to the law of heat conduction, heat flux (power) is proportional to the temperature differential across a solid and is inversely proportional to the thermal insulation properties of the solid (which depend on the thermal conductivity of the material and the thickness, and which are commonly represented by U values). Thus: U_B U_M

The relative thermal conductivity properties of the barrier and of the thermally insulating material of the temperature measuring device may be determined in an initial calibration step, where 9_U is known (e.g. by measuring hot water cylinder outflow temperature, as described below) and θι and θ₂ are measured. Rearranging the equation above,

giving the ratio of insulation between U_B and U_M- This ratio can be used to calculate the temperature 9_U on the remote side of the barrier using the equation:

U_M

Alternatively, if the relative or absolute thermal conductivity properties of the barrier and the thermally insulating material of the temperature measuring device are known, an initial calibration step is not required.

The method thus enables temperature to be assessed remotely or indirectly, enabling determination of the temperature of inaccessible regions.

The insulated cylinder wall is commonly a composite structure, typically having a metal e.g. copper wall with an outer insulating layer.

Given a knowledge of the absolute insulation of the temperature measuring device, U_M, the absolute insulation value of the barrier can be calculated.

The temperature measuring device may be in the form of a heat flux sensor. Heat flux sensors are known per se and are commercially available, e.g. from Hukseflux. Preferably, however, custom devices will be used, typically employing digital IC based sensors, e.g. having an accuracy of +/- 0.5°C or better, tailored to particular requirements, as will be discussed further below.

The temperature sensors are in contact with the thermally insulating material, and are typically spaced apart in a direction extending transversely, e.g. perpendicularly, to the thickness of the barrier i,e, the hot water cylinder wall.

The calibration procedure referred to above can be broken down into the following steps:

1. Heat water in the cylinder fully. It is important to understand that when hot water cylinders are heated from the coil at the bottom, all of the water within the cylinder is heated to the same temperature (ie there is no stratification).

2. By having a temperature sensor on the cylinder outflow pipe, it is possible subsequently to measure the maximum temperature of the water being drawn from the cylinder, which will be the same as the water within the cylinder (0_CW) at the point the cylinder was fully heated. Effectively this is 0_U at the point the cylinder was heated. There could be some reduction in temperature if water is drawn many hours after the heat cycle has ended due to standing losses from the cylinder (ie it has cooled down). This could be easily avoided by aborting any attempt to determine 0_U if no water was drawn within a reasonable time period such as 1 hour.

3. Once we have a valid 0_U value (represented by 0_CW in this case), given that we also have 0₂ and 0i _; as measured by the sensors, we can now calculate the relative value of U_B as compared to U_M as set out above.

4. By knowing the relative insulations of both U_M and U_B and also measuring 0₂ and θ_1; we can at any point now calculate 0_U as explained above. This is the primary concept of this invention, which allows the determination of 0_U in a non-invasive manner.

5. As a secondary concept, if we know the absolute insulation of the temperature measuring device U_M, which is under our control as the sensor designer eg as a U- Value, we can then calculate the absolute insulation value of the cylinder which will provide valuable information for home energy efficiency reports.

To give a concrete example, if we heat a hot water cylinder to 60°C (0_CW, i.e. θ„), and the second temperature sensor temperature is at 20°C (θ₂) and the first temperature sensor reads 40°C (θι) with the system in a quiescent state (ie it has been allowed to settle), we can clearly see that both insulators are of the same insulation levels. If later we then observe θ₂ is 20°C and θι is 30°C, we can extrapolate that the water temperature (0_CW, i.e. θ„) is 40°C, again assuming we are in a quiescent state.

Although after a heat cycle, the cylinder is initially evenly heated, as hot water is drawn from the top of the cylinder and cold water is injected at the bottom, the water becomes stratified with an intersection between hot and cold that moves up the cylinder as more hot water is drawn. Multiple temperature measuring devices can be used at various heights vertically up the side of the cylinder to estimate the level of this intersection, which can then be provided to the home occupant as an indication of how much hot water is available for use, e.g. using hot water management algorithms as disclosed in PCT/GB2011/052281 (WO 2012/069815).

Relative tank levels

The same concept can also be used in a "relative" manner, not requiring an initial calibration step, thus without a sensor on the cylinder outflow pipe. In this case, the goal is not to estimate the absolute temperature of the water (0_CW), but to estimate its relative temperature on a scale of 0 to 100% depending on previous heat cycles. During the learning mode, e.g. carried out by a suitably programmed, remotely located processor to which signals are sent wirelessly, for each heat cycle, θι and θ₂ are monitored, with their extreme limits (highest temperature = hot = 100%, lowest temperature = cold = 0%) learnt over several cycles (eg 3 - 6). After learning by the processor is complete, it is then possible to estimate the relative cylinder level of hot water (0 to 100%)) based on the current temperature readings of θι and θ₂ and the previously learnt limits in historical heat cylinders. This provides an indication of how full the cylinder currently is with hot water as compared to previous heat cycles. This relative method also has the advantage that it is not necessary to know exactly when the heat cycles occurred, as it involves just learning θι and θ₂ limits, so it is ideal for a "monitoring-only" solution with no controllability. The processor may be programmed to recalibrate, e.g. in response to the user changing the tank thermostat temperature. In a simple embodiment, an end consumer use case could involve 3 sensors which the homeowner places on the hot water cylinder at vertically spaced locations, with a simple, convenient remotely located associated processor and display device, eg a wireless LCD fridge magnet, indicating cylinder hot water level. This may provide a low cost embodiment that can be readily retrofitted to an existing cylinder. In this case, the calculated temperature (θ„) is not an absolute value but a relative value. The method of the invention may thus comprise measuring θι and θ₂ over a plurality of cycles of heating of the tank, and calculating therefrom a relative valve of θ„.

Static versus changing readings

When water is consumed from a hot water cylinder, it will take a while before the temperature sensors start to detect this change as they are mounted outside the cylinder insulation. This time delay is mainly dependent on the cylinder insulation material and thickness, and is something that cannot easily be overcome without gaining access directly to the inner cylinder surface skin that borders the water being stored. Delays observed when using typical cylinder insulation vary between 5 minutes and 20 minutes depending on the type of cylinder insulation. If the purpose of knowing the cylinder water temperature is to provide a near-real-time feedback and status display to the home occupants, a delay of this level is acceptable, although it is a draw-back. It is worth noting that when using invasive cylinder sensors (ie cutting away at the cylinder insulation and mounting the sensors directly onto the copper cylinder), this delay is much smaller to the point that it is insignificant.

Once the temperature sensors have started to react to the temperature change, it can take a very long time before they get close to their resting quiescent steady-state, as it roughly follows an exponential decay curve, eventually settling to the resting temperature. It can take over 1 hour before reaching its resting state. By looking at which sensors are changing, and the rate of change, it is possible to make some crude assessments as to how much water has just been drawn from the cylinder.

Specifically:

- If we have installed multiple temperature measuring devices at different heights along the side of the cylinder, we can assess which ones have started to change temperature to provide a very rough estimate of where the new hot water level resides. For example, consider the situation with three temperature measuring devices on the cylinder placed vertically aligned and evenly spaced with the bottom device 1/6 of the height of the cylinder, the middle device half way up the height of the cylinder and the top device 5/6 of the height of the cylinder. If the top device has not started changing, yet the middle device shows rapidly changing temperatures, we know the stratification point separating hot and cold has probably moved above half-way, but has not reached the top device so we know we have between half (50%) and a sixth (17%) of a tank of hot water left.

- By looking at the rate of change of the temperature, a reasonable estimate as to its final resting point can be achieved. In the above example, if the rate of change of the middle device is very small, we are likely to be closer to 50% hot water level. However if the rate of change is rapid, we are probably closer to 17% hot water level. Another example might illustrate the point more clearly. If there is a bigger temperature change within the cylinder (eg from 60°C to 20°C) then when the temperature devices start to change readings, they will change more rapidly than for a smaller temperature change (eg 60°C to 58°C). This allows for a crude estimate of where the temperature is heading to.

More precise corrections can be made, as discussed below in the "Calculation corrections" section.

By observing the history of previous heating and cooling cycles, the system can be further optimised to improve accuracy. For example, a certain rate of change typically resulted in a specific condition, eg an initial decrease of 0.2°C per minute from the first temperature sensor may have resulted in a 10°C water temperature decrease after quiescent state was achieved (perhaps 15 minutes later). We can now take this historical information into account to attempt to predict more accurately the final resting temperature given a current rate of change. This maps well to consumer hot water consumption, as water is typically drawn in short bursts rather than continually so a short burst of hot water activity would soon result in rapid rate of change, allowing for an early estimate of the current tank temperature even before quiescent state has been achieved. Errors

The individual temperature sensors will carry an error margin, and which typically be +/- 0.5°C or less depending on the quality and manufacturer of the sensors used.

When performing calculations on extrapolated sensor readings, sensor errors are magnified depending on the level of extrapolation. This means that the temperature measuring device insulation needs (U_M) to be "significant" in comparison to the barrier insulation level (U_B) to keep the error magnification minimised. The error magnification factor is 1+ (U_B / U_M).

So for instance, if the sensors have an accuracy of +/-0.5°C, and the insulations are both the same (ie U_M = U_B) the error magnification factor is 2, providing a final extrapolated temperature accuracy of +/-1.0°C. Using more accurate sensors enables error levels to be reduced.

Depending on the overall system error requirements, the temperature measuring device insulation level can be chosen to meet those objectives. In the case of providing a visual hot water level feedback to the home occupants, an inaccuracy in water temperature of 2°C or 3°C is perfectly acceptable, and assuming native sensors error is less than 0.5°C this would indicate that the temperature measuring device insulation must be at least 1/4 the level of the cylinder insulation or better. A foam skinned hot water cylinder has a typical U value in the range 1 - 2 W/m²K, with a double skinned metal cylinder having a typical U value of about 0.7 W/m²K, so appropriate U values for the temperature measuring device insulation can be calculated accordingly. Lower U values represent greater levels of insulation.

Further errors will be introduced if the system is not in a steady state, and there will likely have a bigger impact of accuracy. However quantifying these errors is not so easy, as it depends on the time constant of the temperature decay curve to its quiescent state. Initially these errors will be significant, but will reduce as temperatures approach quiescent state. All of the above errors indicate that this technology is not suitable for accurate temperature estimation of water within the cylinder, however it is perfectly fit for purpose when attempting to assess how much hot water is available to the home occupants.

Calculation corrections

The calculations above assume that temperatures can be extrapolated according to a simple ratio of insulation values, but this is only true under a number of assumptions. Specifically, a) the system temperatures must be constant and b) the energy flow from the cylinder to the inner sensor must be the same as the energy flow from the inner sensor to the outer sensor.

Condition a) is violated when cylinder temperatures are changing rapidly (e.g. during a heating cycle or when hot water is consumed and replaced by cold water). In this case, and outer temperatures are out of sync with changes reaching the inner sensor first, and so the extrapolation no longer works accurately. This can be corrected for, either by estimates as described above in the section "Static versus changing readings", or by modelling heat propagation or dissipation through the insulating material. Note that no method can eliminate the delay in reading cylinder level; these methods merely reduce the error in estimating the level during periods of change.

Condition b) is violated for example when there is insufficient area of insulation around the non-invasive sensors, or when geometry constrains the sensors to be closer together in relation to the insulation thickness. A simple correction can be made by assuming a proportion of heat transfers to the neighbouring sensor, and solving the resulting simultaneous linear equations. This also allows for other sensor geometries, such as secondary sensors midway between the primary sensors, or a greater density of sensors (for greater accuracy). Thus, correction can be made for shared heat flux between sensors by estimating the coupling and solving the resulting linear equations.

In general, the accuracy of the method can be improved through the use of a thermal model for the temperature distribution of the water in the cylinder, and how this changes with time and propagates through the insulating materials. We can then infer the temperature profile inside the cylinder that best matches the temperature measurements we have made, and also estimate how long ago this profile was true (which can be used to guide hot water management, for example firing for hot water earlier to reduce the likelihood of the cylinder running out).

Hot water cylinder and temperature measuring device construction

The invention also provides a hot water cylinder, having fitted to the outer surface of an insulated wall of the cylinder at least one temperature measuring device comprising a first temperature sensor in contact with the outer surface of the cylinder wall and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material.

Preferably two, three or more separate temperature measuring devices are fitted to the outer surface of the cylinder wall of the cylinder at vertically spaced locations, eg being vertically aligned at approximately 1/6, ½ and 5/6 of the height of the cylinder, to enable assessment of hot water temperature at different levels in the cylinder.

The or each temperature measuring device is conveniently connected (by wires or wirelessly) to a processor programmed to calculate water temperature from signals supplied thereto. Display means, separate or combined with the processor, are desirably provided for displaying calculated temperatures.

The temperature measuring device preferably uses thermally insulating material that has a curvature to match that of the hot water cylinder wall. Yet more preferably, the thermally insulating material is flexible to allow the device to be used with cylinders of different radius, shape, etc. As noted above, the thermally insulating material should have good insulating properties to reduce error magnification factors. For ease of installation and manufacturability and to reduce response times, the thermally insulating material is desirably relatively thin. It is thus preferred to use a flexible material with good insulation properties, e.g. having a thermal conductivity of 0.02W/mK or less. Particularly suitable materials include sheets of fibrous matting impregnated with silica aerogel, e.g. in the form of Spacetherm (Spacetherm is a Trade Mark) blankets, e.g. of 10 mm thickness. Spacetherm has a thermal conductivity of 0.013 W/mK so a 10 mm thickness has a U value of 1.3 W/m²K. This is approximately twice the U value of a typical double skinned metal tank (about 0.7 W/m²K), and so represents about half the level of insulation, and so satisfies the error requirement discussed above.

The temperature measuring device should be significantly wider than the thickness of the cylinder insulation in order to get an accurate reading of heat flow, otherwise heat flowing through the cylinder insulation will flow around the flux sensor rather than through it. A 10mm thick circular disc with a diameter in the range 15 to 20cm with temperature sensors placed in the centre of each major side is suitable to provide sufficiently accurate results for a typical hot water cylinder. It is important that the device is bonded to the cylinder to avoid air gaps or air flow. The second temperature sensor measuring the exposed outer surface temperature will react to ambient air temperature changes much more quickly than the first temperature sensor and may benefit from software "dampening" such as a software window averaging algorithm over the last 30 minutes. Physical dampening could also be implemented by adding an extra small layer of insulation on the second outer sensor.

In a further aspect, the invention provides a temperature measuring device for use with a hot water cylinder, comprising a sheet of flexible thermally insulating material having a first temperature sensor on one face thereof and a second temperature sensor on the opposed face thereof.

The insulating material preferably has a thermal conductivity of 0.02 W/mK or less and a U value of 4 W/m²K or less, more preferably 2 W/m²K or less, yet more preferably 1.5 W/m²K or less.

The insulating material conveniently comprises a sheet of fibrous matting impregnated with silica aerogel.

The insulating material desirably comprises a disc with a diameter 15 to 20 cm, with a respective temperature sensor in the centre of each of the major faces.

The invention may be put into effect by retrofitting suitable temperature measuring devices to an existing hot water cylinder. Thus the invention also provides a kit for use with a hot water cylinder, comprising at least one temperature measuring device comprising a first temperature sensor and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material; a processor programmed to calculate water temperature (absolute or relative) from signals supplied thereto by the temperature measuring device(s); and display means for displaying the calculated temperature(s).

The kit preferably also includes a temperature sensor for the cylinder outflow pipe, arranged to supply signals to the processor.

The kit desirably also includes instructions for use.

The kit preferably includes three temperature measuring devices.

The temperature measuring devices are preferably as discussed above.

The kit may include connecting wiring for connecting the temperature sensors to the processor, or wireless connections may be employed.

The invention will be further described, by way of illustration, with reference to the accompanying drawings, in which:

Figure 1 is a schematic sectional view of part of the wall of a hot water cylinder, with a temperature measuring device mounted thereon; and

Figure 2 is a schematic side view of a hot water cylinder fitted with a kit in accordance with the invention.

Detailed description of the drawings

Figure 1 illustrates schematically a sectional view of part of a barrier in the form of the composite wall 10 of a hot water cylinder having a remote (inner) side 12 and a near (outer) side 14. The composite wall 10 comprises a copper wall 16 with the remote side in contact with water 18 at a temperature of 9_CW (i.e. 9_U) with an outer layer 20 of insulating material having thermal conductivity properties U_B. A temperature measuring device 22 is mounted on the near side 14 of the composite wall 10. The device 22 comprises a first or inner temperature sensor 24 and a second or outer temperature sensor 26, with both sensors enclosed within a thermally insulating material 28 having thermal conductivity properties U_M-

Figure 2 shows a hot water cylinder 40 with three temperature measuring devices 42, 44, 46 secured to the outer surface of the cylinder wall in vertical alignment, at approximately 1/6, ½ and 5/6 of the height of the cylinder. The devices are connected via wires 50 to a processor unit 52 including display means. A temperature sensor 54 is secured to the cylinder out flow pipe 56, and is connected to the processor unit 52 by wire 58.

Each temperature measuring device 42,44,46 is of similar construction and comprises a 20 cm diameter circular disc of 10 mm thick Spacetherm flexible aerogel blanket having an adhesive backing with removeable covering. The material has a thermal conductivity of 0.013 W/mK of the disc has a U value of 1.3 W/m²K. A respective temperature sensor is placed in the centre of each major face of the disc, without air gaps. The sensors are each in the form of a TMP275 digital IC based sensor from Texas Instruments, having an accuracy of about +/- 0.2°C over the temperature range of interest, A small piece of thermally insulating material is located on the sensor on the outer, non-adhesive side of the disc. The removable covering of the disc is removed to expose the adhesive backing of the sensors, which secured to the side of the cylinder, being careful to avoid any air gaps, with the blanket readily conforming to the shape of the cylinder.

The temperature measuring devices 42, 44, 46 are used as described above to assess remotely the temperature of water inside the cylinder at the level of the sensor, after an initial calibration step as described.

List of Symbols θι = temperature at the first temperature sensor

θ₂ = temperature at the second temperature sensor

9_U = unknown temperature on the remote side of the barrier U_B = thermal conductivity properties of the barrier

U_m = thermal conductivity properties of the thermally insulating material of the temperature measuring device

6_H = rate of heat flow

9_CW = temperature of cylinder water

Claims

1. A method of assessing the temperature on the remote side of a barrier in the form of an insulated wall of a hot water cylinder, comprising applying to the near side of the barrier a temperature measuring device comprising a first temperature sensor in contact with the near side of the barrier and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material, measuring the temperature (θι) at the first temperature sensor and the temperature (θ₂) at the second temperature sensor, and calculating from the measured temperature values the temperature (9_U) on the remote side of the barrier, based on the thermal conductivity properties of the barrier (U_B) and of the thermally insulating material of the temperature measuring device (U_M).

2. A method according to claim 1, comprising measuring θ_1; and θ₂ when 9_U is at a known value to determine the relative thermal conductivity of the barrier and of the thermally insulating material in an initial calibration step.

3. A method according to claim 2, wherein 9_U is determined by measuring the temperature of water drawn from the cylinder after fully heating water in the cylinder.

4. A method according to claim 3, comprising assessing the temperature at two or more vertically spaced locations on the hot water cylinder wall using respective temperature measuring devices.

5. A method according to claim 4, comprising monitoring relative rates of temperature change at the vertically spaced measuring devices, and calculating therefrom the estimated proportion of hot water in the cylinder.

6. A method according to any one of the preceding claims, wherein the thermal insulation properties of the insulating material of the temperature measuring device are at least ¼ of the level of the barrier insulation.

7. A method according to any one of the preceding claims, further comprising calculating the absolute insulation value of the barrier, based on a knowledge of the absolute insulation value of the insulating material of the temperature measuring device.

8. A method according to any one of the preceding claims, comprising measuring θι and θ₂ over a plurality of cycles of heating of the remote side of the barrier, and calculating therefrom a relative valve of 9_U.

9. A hot water cylinder, having fitted to the outer surface of an insulated wall of the cylinder at least one temperature measuring device comprising a first temperature sensor in contact with the outer surface of the cylinder wall and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material.

10. A hot water cylinder according to claim 9, wherein at least two separate temperature measuring devices are fitted to the outer surface of the cylinder wall at vertically spaced locations.

11. A hot water cylinder according to claim 10, wherein respective temperature measuring devices are fitted to the outer surface of the cylinder wall at approximately 1/6, 1/2 and 5/6 of the height of the cylinder.

12. A hot water cylinder according to any one of claims 9 to 11, wherein the or each temperature measuring device is connected to a processor programmed to calculate water temperature from signals supplied thereto.

13. A hot water cylinder according to any one of claims 9 to 12, where the or each temperature measuring device comprises a sheet of flexible thermally insulating material.

14. A hot water cylinder according to claim 13, wherein the flexible thermally insulating material has a thermal conductivity of 0.02 W/mK or less.

15. A hot water cylinder according to claim 13 or 14, wherein the flexible thermally insulating material has a U value of 4 W/m²K or less.

16. A hot water cylinder according to claim 13, 14 or 15, wherein the flexible thermally insulating material comprises a sheet of fibrous matting impregnated with silica aerogel.

17. A hot water cylinder according to any of claims 13 to 16, wherein the flexible thermally insulating material comprises a disc with a diameter in the range 15 to 20 cm.

18. A temperature measuring device for use with a hot water cylinder, comprising a sheet of flexible thermally insulating material having a first temperature sensor on one face thereof and a second temperature sensor on the opposed face thereof.

19. A device according to claim 18, wherein the insulating material has a thermal conductivity of 0.02 W/mK or less.

20. A device according to claim 18 or 19, wherein the insulating material had a U valve of 4 W/m²K or less.

21. A device according to claim 18, 19 or 20, wherein the insulating material comprises a sheet of fibrous matting impregnated with silica aerogel.

22. A device according to any one of claims 18 to 21, wherein the insulating material comprises a disc with a diameter in the range 15 to 20 cm with a respective temperature sensor in the centre of each of the major faces.

23. A kit for use with a hot water cylinder, comprising at least one temperature measuring device comprising a first temperature sensor and a second temperature sensor spaced from the first temperature sensor and separated therefrom by thermally insulating material; a processor programmed to calculate water temperature from signals supplied thereto by the temperature measuring device(s); and display means for displaying the calculated temperature(s).

24. A kit according to claim 23, further comprising a temperature sensor for the cylinder outflow pipe.

25. A kit according to claim 23 or 24, further comprising instructions for use.

26. A kit according to claim 23, 24 or 25, including three temperature measuring devices.

27. A kit according to any one of claims 23 to 26, wherein the or each temperature measuring device is in accordance with any one of claims 18 to 20.

28. A method of assessing the temperature on the remote side of a barrier, particularly a hot water cylinder, substantially as herein described with reference to the accompanying drawings.

29. A hot water cylinder substantially as herein described with reference to, and as shown in, the accompanying drawings.

30. A temperature measuring device substantially as herein described with reference to, and as shown in, the accompanying drawings.

31. A kit for use with a hot water cylinder, substantially as herein described with reference to, and as shown in, the accompanying drawings.