• No results found

Chapter 3: Experimental Performance of BIT Collector and SWH System

3.3.2 Standard Load Profile Performance

It has been stated that an experiment using the BIT-SWH to investigate the performance of the collector would be meaningless, however from the viewpoint of the control system; such a test may provide interesting results. Therefore it was decided that an investigation into the effectiveness of the controller for use in a BIT-SWH would be undertaken.

It should be noted that the particular day this test was performed is not considered

‘optimum’ solar conditions, with an average solar insolation of 126 W/m2 for the period where the test was performed between 8 am to 8 pm. The results for the tank temperature distribution are shown in Figure 41. Additionally, the collector

Figure 41: Storage tank temperature distribution for standardised load.

Figure 42: Collector temperature and environmental readings.

23.4 23.6 23.8 24.0 24.2 24.4 24.6 24.8 25.0

6:00 a.m.8:24 a.m.10:48 a.m.1:12 p.m.3:36 p.m.6:00 p.m.8:24 p.m.10:48 p.m.

Temperature C)

Local Time

Tt1 Tt2 Tt3 Ttank Telement

19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5

0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0

6:00 a.m.8:24 a.m.10:48 a.m.1:12 p.m.3:36 p.m.6:00 p.m.8:24 p.m.10:48 p.m.

Temperature C)

Radiation (W/m2)

Local Time

Radiation Ambient Collector Inlet Collector Outlet

The controlling of the system pump and solenoid were based on the differential principle and standard load profile, respectively. It was observed that for the conditions of the testing period the parameters for meeting the algorithm for the pump were not fulfilled. The temperature difference needed to activate the pump was set at 8 °C; however as can be seen from Figure 41 and Figure 42, the values which dictate this difference (being collector outlet and tank cold region temperature, Ttank) doesn’t actually produce a temperature difference. This is because the collector did not receive sufficient heat gains due to relatively low solar insolation. Similarly, the solenoid was never activated; this is largely due to the fact that there wasn’t sufficient heat gain in the storage tank to supply to the load. Initially, the algorithm was configured to ‘dump’ different volumes of water so as to represent the load used by the household however; this was impractical due to the fact that the load needed is based on hourly requirements in terms of thermal energy (MJ/hr). Therefore the controller was reconfigured to dump a certain amount of energy at times specified by the standard profile. Because of the direct relation between the collector heat again and thermal energy applied to the tank, the signal to the solenoid valve remained closed. Details of the algorithm are provided in Appendix D.

Since the performance of the controller could not be determined using the experimental rig, a simplified theoretical version for the pump and solenoid, shown in Figure 43 and Figure 44 respectively, was developed allowing different cases to be applied. This would allow the effectiveness of the controller for use in a system such as the BIT-SWH to be investigated.

Figure 43: Simplified pump controller (left: front panel, right: block diagram).

Figure 44: Simplified solenoid controller (top: front panel, bottom: block diagram).

For the pump the following cases of activation/deactivation were investigated:

Temperature difference (ΔT) over 8 °C, below 4 °C and worst case scenario where the temperature in the hottest region of the tank (Tt1) exceeds 90 °C. The results from the theoretical controller showed promising results, with the pump (LED light) activating when the ΔT was greater than 8 °C and deactivated when the ΔT was less than 4 °C. The pump would also remain deactivated until the ΔT of 8 °C was reached again. This means that the pump would only activate if there was heat to be gained from the collector. Additionally, the pump remained deactivated for the worst case scenario when the temperature in the hottest region, Tt1, exceeded 90 °C even if there was heat to be gained from the collector. In contrast, the load controller activation is based on the current local time rather than parameters from the system. All the cases set out in the load profile, as shown in Table 1 were investigated.

Table 1: Load profile cases

Time Hourly thermal load (MJ/hr)

6:00 a.m. 1.13

8:00 a.m. 3.62

10:00 a.m. 3.39

12:00 a.m. 2.49

2:00 p.m. 1.81

4:00 p.m. 1.58

6:00 p.m. 2.49

8:00 p.m. 2.94

10:00 p.m. 2.49

Once the parameter for activation was met (the local time) load was drawn from the system, for example at 6:00 am a required load of 1.13 MJ/hr needs to be drawn. This was achieved by actively comparing the load value of 1.13 MJ/hr to the rate of heat extracted, Qtank, from the tank, calculated using Equation 23.

When the load matched the extracted energy, the valve would then shut. This would then be repeated for all time periods, thus replicating the standardised load profile.

Qtank = m Cp (Texit – Tmain) (23)

Where, m is the flow rate through the tank in kg/s and Texit and Tmain are the outlet and inlet temperatures, respectively, in °C.

For the tests, an approximate flow rate of 0.17 kg/s (10 L/min) was used; this was obtained by manually measuring the volume of mass exiting the tank within 60 seconds. Additionally, the inlet temperature was taken to be 10 °C and outlet as 60

°C which was manually decreased over time to simulate the cooling of the water in the tank as load was drawn.

The results from the theoretical controller indicated that this method of simulating the load for a household maybe impractical. For instance, if the highest and lowest load cases are considered, (8:00 am with an hourly load of 3.62 MJ/hr load and

on heat extracted from the tank can be calculated. Of the parameters in Equation 5, the temperature difference is the only parameter which varies (assuming constant flow rate). The results for the two extreme cases are summarised below (for all time periods refer to Appendix E).

8 am using 3.62 MJ/hr, ΔT = 1.4 °C required 11 pm using 0.68 MJ/hr, ΔT = 0.3 °C required

It can be seen that for the highest heat load, a temperature difference of 1.4 °C is needed before the amount of extracted heat from the matches that of the load.

Moreover, for the lowest heat load, a temperature difference as low as 0.3 °C is required. This highlights a slight impracticality in the method. It is important to note however, that a big assumption made here is that the heat gain from the collector doesn’t affect the temperatures within the tank. This would be untrue in an actual application. Unfortunately, this could not be investigated further with this system, due to time constraints and the collector being classified as ‘poorly’

performing. However, this exercise has shown that the controller is accurate and can be effective. What is also advantageous about the controller is that if perhaps a different approach was used compared to the method used here, a slight modification of the controller can be made to accommodate for this. With this mind, different aspects such as incorporating live weather data and automatic prediction of user water draw off (Prud’homme and Gillet, 2001) by implementing advanced control strategies. This would ultimately result in a well-managed, efficient system.

3.4 Experimental Summary

From the experimental testing of a BIT collector it was shown that an optical efficiency of 75% could be achieved, the downside to this is that high heat losses also occurred. It was identified that the major contributing factor to the heat loss was the lack of side insulation as well as air spaces present between the collector and the polystyrene insulation. Moreover, a possible solution was suggested for overcoming this by using extruded polystyrene sheets with a matching profile, thus eliminating the spaces of air and also extending to provide thermal resistance on the side. From a theoretical viewpoint of the glazed collector it was shown that the major contributions to the heat gain and heat loss of the system was the fin efficiency and overall heat loss coefficient, particularly the contribution of the bottom and side heat loss coefficients. Moreover, it was shown that the use of a higher thermal conductivity material does in fact increase the fin efficiency and subsequently the heat gain and thermal efficiency of the collector.

Additionally, it was decided that rather than optimise the BIT collector, experiments would be conducted to investigate its suitability in a domestic water heating application. The reason for this was largely due to the fact that the BIT collector was designed to replicate the roofing structure of a standard metal long run roof, which is typical of New Zealand house structures.

Results from the systems test when no load was drawn showed that temperatures of over 60 °C could be reached within the tank. A comparison with previous

highlighted the potential pitfall of using a simple design, which would also suggest that the high heat loss from such a design has consequently adverse effects on it effectiveness for water heating. As such, it was decided that it would be impractical to investigate its use in a system replicating daily operation.

Another aspect of the BIT-SWH that was of interest was the control system. This is used for controlling both the pump and solenoid valve using temperature differential control and a standardised load profile. Although it was decided that it was impractical to run a replicate SWH system to characterise collector performance, from the viewpoint of the control system such as test was of interest.

Unfortunately, the experimental tests in less than ‘ideal’ solar conditions failed to activate either of the controllers; therefore a theoretical approach was used to investigate its effectiveness.

From the theoretical tests conducted on the controllers, it was seen that the algorithm is indeed effective for controlling a BIT-SWH system. This is especially true for the temperature differential pump controller. Tests also showed that the ‘dumping’ controller was accurate but perhaps a different method to the one used would be more practical. Additionally, it was identified that there was potential to modify the collector to a more advanced strategy.