• No results found

Chapter 1: Introduction

1.5.2 Building Integrated Thermal Collectors

In summary the majority of works on flat plate collectors have focused on improving heat transfer characteristics through the optimisation of different components, reducing heat losses within different designs and development of new and innovative designs which go hand in hand with the development of technology in the area of solar heating.

This view was also shared by Medved et al. (2003), who examined an unglazed

‘solar thermal’ system, essentially a BIT collector, that could be truly integrated into a building, as shown in Figure 5. In their system they utilised standard metal roofing as a BIT collector for the heating needs of an indoor swimming pool.

They found that they were able to achieve payback periods of less than 2 years.

This translated to a reduction of 75% in the time taken to pay for a glazed system.

Furthermore, they report that they were able to achieve efficiencies of over 80%

however; these were achieved using collectors with areas of over 200 m2.

Figure 5: Unglazed integrated solar collector (adapted from Medved, et al., 2003).

Assoa et al. (2007) also presented a concept for a roof integrated water and air heating collector with a high level of integration. They designed a collector in which a water tube was placed in the trough of a troughed-roof system with PV cells added to the ridges, as shown in Figure 6. This type of collector combines

dimensional mathematical model of their systems. Using a parametric study (numerically and experimentally) the effect of various factors were investigated, such as the mass flow rate of the water on the solar collectors’ thermal performance. They note that the water solar collector mass flow rate has very little influence on the solar air collector behaviour. Furthermore, they theoretically demonstrated that combined efficiencies in excess of 80% could be achieved by their system.

Figure 6: Water and air heating roof integrated collector (adapted from Assoa, et al., 2007).

A similar design to that of Assoa, et al. (2007) was also presented by Anderson et al. (2009). They designed and developed a roof integrated prototype collector that was integrated into a standing seam or troughed sheet roof, along with passageways for liquid coolant flow, as shown in Figure 7. Using a modified Hottel–Whillier model they validated results using an outdoor steady-state thermal test rig. Their results concluded that a few key design parameters such as

the PV module and the supporting structure have significant effects on both the thermal and electrical efficiencies. On the thermal side, they report that they were able to achieve efficiencies of 60% and 30% for glazed and unglazed, respectively. In a concluding remark they suggested that the integration of systems ‘into’ (rather than ‘onto’) the roof structure did not need insulation, as the rear air space in the attic can provide a level of insulation equivalent to a highly insulating material.

Figure 7: BIPVT collector design (adapted from Anderson, et al., 2009).

Similarly, Wahab et al. (2011) investigated a BIT collector for water heating.

They investigated an identical design to that of Anderson, et al. (2009), by using standard long run metal roofing material to produce an integrated (BIT) collector, as shown in Figure 8.

Figure 8: Long run metal roofing BIT collector (adapted from Wahab, et al., 2011).

The authors report that with their collector, they were able to heat an insulated tank (~35 L) of water to approximately 90 °C on a clear sunny day (average solar insolation of 929 W/m2). Additionally, they report a thermal efficiency of 45 % for the glazed BIT collector and also stated that although the efficiency was relatively low; their system would still be effective as a solar water heater in

‘sunny’ regions (referring to Australia and New Zealand). Furthermore, they highlight that the possibility of integrating an ‘effective’ solar hot water system directly into standard roofing material is viable, thus maintaining the aesthetics of the building. Moreover, the main focus of their work was to develop an

‘advanced’ control strategy for the control of their system, as such, the work from this research was only to serve as a pre-requisite for future work based on control.

1.6 Hypothesis

In light of the review on previous literature, particularly on SWH systems, there is still room for work to be undertaken particularly to ‘experimentally’ investigate the performance of building integrated SWH systems, specifically in-situ systems with reference to New Zealand conditions. It was observed that the majority of previous research on SWH systems was conducted on commercially packaged systems, which were optimised for flat plate and evacuated tube collectors.

Additionally, it has been identified that the majority of solar thermal systems in New Zealand are predominantly water heating systems in the domestic sector (Weiss and Mauthner, 2012).

While previous works on BIT collectors have addressed to some extent the issue of integration and architectural uniformity, there is still room for improvement in terms of achieving both high levels of integration and adequate collector performance. It has been shown that unglazed collectors such as that presented by Medved, et al. (2003) can achieve very high peak efficiencies (>80%), the downside to this is that although it is being used in a low temperature heating application (pool heating) large areas (>200 m2) are required to achieve effective pool heating. Furthermore, the lack of glazing and insulations makes it unsuitable for domestic water tank heating as it can rarely attain the required temperatures. In an application such as SWH where roofing space is limited, consideration should be made to accommodate for this when designing and developing the collectors.

Although the collectors studied by Assoa, et al. (2007) and Anderson, et al. (2009) have shown good performance, both have incorporated PV cells, which to some extent adversely effects the performance of the overall collector when considering the thermal side. This is because it is desirable to maintain relatively low cell temperatures to achieve optimum cell performance.

One study (Anderson, 2009) did however, investigate the long term performance of their collector (Building integrated photovoltaic/thermal, BIPVT) in a SWH system applicable to New Zealand, but was limited to a simulation analysis using the TRNSYS software and typical meteorological year (TMY) data. The work by Wahab, et al. (2011) provided an indication of how such systems could perform when in a domestic setting, however, the volume of water tested (~35 L) does not represent the volume of consumption of a typical household. Issacs et al. (2010) have found that for showers alone, assuming an average flow rate of 8.4 L/min the average water consumption is 200 L per day.

Additionally, the work (Wahab et. Al, 2011) was only to serve as a prerequisite for their future work on the control of BIPVT systems. To the author’s knowledge there is currently little published work that focuses ‘specifically’ on the performance of ‘truly’ BIT collectors for use in SWH systems. For this statement conventional flat plate thermal collectors that are flush with the surface of the building façade or roof are not considered ‘truly’ building integrated.

On this premise it can be said that there still a need for the development of BIT collectors which achieve a high level of integration as well as good collector performance. As a result an experimental investigation of the performance of BIT collectors in an application such as SWH can be undertaken to determine its applicability in a domestic setting, for a country like New Zealand. Therefore the aims of this thesis are to:

 develop a truly integrated BIT collector which addresses architectural uniformity while achieving good collector performance

 investigate the practicality of using the BIT collector in SWH systems for domestic application in New Zealand

1.7 Methodology

Different methodologies exist for the evaluation of the hypothesis described above. It was observed that there is room for improvement in terms of the designing of building integrated collectors to achieve good performance as well as maintaining the building aesthetics. In this case an attempt can be made to design a collector replicating that of a metal long run roof, which is typical of a roofing structure in New Zealand. The reason for this is that such a design could potentially be integrated ‘into rather than onto’ the roofing structure as suggested by Anderson (2009).

Additionally, different methodologies are available for evaluating and characterising the performance of such collectors. It has been observed from previous literature that there is a need for experimental validation of such collectors and perhaps the same evaluation can be applied for a system running with the collector. It has been stated that currently there is no set configuration for BIT water heating systems and therefore its suitability in a domestic setting is yet to be known. Therefore, a combined simulation and experimental approach for such systems can provide a way of assessing and validating its performance.

Based on these premises a methodology was selected to design a BIT to replicate metal long run roofing, experimentally characterise its performance and approach the system performance evaluation using simulation and experimental work.

Chapter 2: Development of a Building Integrated Thermal Collector and Building Integrated Thermal-Solar Water

Heating System

2.1 Introduction

The following chapter describes the development of a new design of a BIT collector including the test procedures undertaken to characterise its performance, in terms of efficiency. The subsequent result of this development is that the applicability of the collector for SWH can be investigated. The remaining sections within this chapter describe the development and testing procedure of the BIT-SWH system

2.2 Development of a BIT Collector

The BIT collector used in the SWH system in this study is unique in a number of ways. Similar to the collector design developed by Anderson (2009), this collector can be directly integrated into the roof of a building thus providing a high level of integration. Standing seam and troughed sheet roofs are typically made from aluminium or coated steel, although copper or stainless steel could be used.

The choice of aluminium as the base material for the collector provides high thermal conductivity among other things. Using aluminium, the panel is extruded into a roof-profile shape that gives the roof product stiffness and strength, and when assembled are weather proof. The choice of material provides several benefits. Aluminium is a very versatile material with a range of advantageous properties (Richards, 2009):

 Lightweight - is about one-third the weight of an equal volume of copper, steel or brass. This is especially important as this reduces the load applied on the roofing structure

 Strength - Aluminium can withstand heavy loads and pressure;

when alloyed appropriately, its strength approaches that of steel.

As a result it has a high strength-to-weight ratio.

 Corrosion resistant - aluminium oxide on the surface of the metal protects it against the corrosive influences of water, salt and other influences.

 Good thermal conductivity - Aluminium distributes heat or cooling energy evenly and quickly.

 Ductile - Aluminium is easy to cold work and fabricate.

 Malleable - The vulnerability of ‘pure’ aluminium to heat and pressure make it ideal for extruding into formed, intricate shapes.

Figure 9: BIT roof panel profile.

Furthermore, the manufacturing process of extruding aluminium is a well-established process (Bauser and Siegert, 2006). One of the many advantages of producing extrusions is the ability to create profiles with intricate shapes and patterns that would unless otherwise be manufactured separately using other manufacturing processes (Bauser and Siegert, 2006). An important component that can be directly produced through extrusion is the channels required for fluid flow, shown in Figure 9; this eliminates the need for further manufacturing, therefore reducing the overall cost of the collector. Another unique feature is the

‘interlocking’ nature of the extruded panels which provides the opportunity to assemble the collector to a desired geometry, as shown in Figure 10. The aluminium panels were also anodised to increase corrosion resistance (Richards, 2009) and dyed black to give a good surface finish while providing good absorption over the solar radiation spectrum (Anderson, 2009).

Figure 10: Overlapping panels, creating a BIT collector.

Due to the simple nature of the BIT panels the construction of the collectors was relatively straightforward. The panels were interlocked together to create the desired collector area and mounted onto the testing frame, as shown in Figure 11.

The ends of the channels which provide the inlet and outlet (risers) for fluid flow were tapped and fitted with hose fittings so that the manifold could be attached.

Based on a numerical analysis specific to this collector design, a manifold to riser pipe ratio of 4:1 was selected as this was shown to be the ideal configuration for achieving a uniform flow distribution throughout the collector (Ghani et al., 2012). The collector risers were linked to the copper manifold (header) using flexible wire (f-w) hose and clamped using standard hose clamps, shown in Figure 12. It should be noted that this method of using f-w hose would not be suitable for a commercial BIT due its unpredictable behaviour under relatively high temperature operation over very long periods, but this was considered adequate for research purposes.

Figure 11: Interlocked panels forming an unglazed collector.

Figure 12: F-w hose clamped onto manifold.

For both the glazed and unglazed BIT collectors the ends of the roof profile were enclosed and the rear surfaces insulated with 50 mm expanded polystyrene insulation, as shown in Figure 13. The glazing used was a low-iron-glass cover placed over the collector to prevent convection losses, side flashings were also used thus forming a glazed building integrated collector with a roof profile, as shown in Figure 14. It should be noted here that the overall collector design was configured to replicate the simplest and most practical configuration of a typical New Zealand household roof, as shown in Figure 15. The design parameters of the BIT collector tested are provided in Appendix A.

Figure 13: End caps on top and bottom of BIT collector with 50 mm polystyrene insulation on underside.

Figure 14: Glazed BIT collector with side flashings and top and bottom end caps.

Figure 15: Typical profile of a metal long run roof used in New Zealand.

2.3 Collector Testing Setup

The task of investigating collector performance has been an exercise undertaken for many years. With focus on collector testing in New Zealand conditions, standards such as AS/NZS 2535.1-2007 and ISO 9806-3 can be applied. For this study a steady state outdoor thermal test setup was used, similar to that recommended in AS/NZS 2535.1-2007 and shown in Figure 16 (a schematic diagram is also provided in Appendix B).

Figure 16: Steady state solar collector testing system.

The main reason for using an outdoor testing system is that the collector performance is characterised based on an actual solar spectrum (Anderson, 2009).

This method of testing also provides an insight into conditions expected at a location within New Zealand (Hamilton). Experiments conducted by Anderson

testing collectors using indoor setups, the need for specialised equipment makes this process financially unviable for this study. Indoors tests are generally conducted using solar simulators, “that is, a source producing radiant energy that has spectral distribution, intensity, uniformity in intensity and direction closely resembling that of solar radiation” (Duffie and Beckman, 2006). The results are also not always comparable to those of outdoor tests because the diffuse-fraction and long wave radiation (which are not the same indoors and outdoors) can affect the relative results of tests (Gillett, 1980). Other studies (refer to Garg et al., 1985, Tiedemann and Maytrott, 1997 and Adelhelm and Berger, 2003) have also found that using indoor solar simulators showed either poor replication of the solar spectrum or non-uniform illumination, further supporting the use of an outdoor system.

To obtain accurate test results the test location is important, therefore an unimpeded north facing test location at The University of Waikato’s Aquatic Research Centre was chosen, Figure 17 (left). To quantify the performance of the collector it was necessary to measure the global incident solar radiation at the test location. The measurement of the incoming radiation was made using a WMO First Class pyranometer which had been recently calibrated, mounted in-line with the collector at an angle of 38 °C which is equal to the local latitude, shown in Figure 17 (right). A cup anemometer which is used to monitor wind speed in the test area was mounted adjacent to the test stand for the collector, shown in Figure 17 (right).

Figure 17: Left-unimpeded test location at the Aquatic Centre, right-pyranometer and cup anemometer.

K-type thermocouples were used to measure the inlet and outlet temperatures to the collector and the local ambient air temperature, Figure 18 (left). In addition to the measurement apparatus, an instantaneous gas water heater with an inbuilt temperature controller was also mounted on the inlet side of a 700 litre tank, as shown in Figure 18 (right). The outlets from the collector and the water heater was returned to the water tank, as shown in Figure 19, the reason for this was to provide a means of controlling the inlet water temperature which for fills the requirements for collector testing as set out in the AS/NZ 2535 standard.

Figure 18: Left-thermocouples mounted to inlet and outlet, right-instantaneous water heater and water tank.

Figure 19: Fluid channels to and from the collector and water heater.

The flow of water through the collector was set at a constant rate by manually adjusting the valves at the inlet to the collector, shown in Figure 19. The flow rate values were measured by manually measuring the time taken for a known mass to pass through the collector.

The testing of the BIT collectors was conducted in accordance with AS/NZS 2535.1-2007. This standard specifies a test method to determine the thermal efficiency of solar collectors. A prerequisite to accurately determining the performance of the collector is to conduct a number of outdoor tests under a range of ambient conditions. In this study both a glazed and unglazed BIT collector were tested.

To collect all the data needed for analysis, a data acquisition (DAQ) program was developed using National Instruments (NI) systems design software, labVIEW, shown in Figure 20. This program allowed the user to control all aspects of data logging including, save file locations, tested collector area and flow rate and user specified logging intervals.

For each test the temperatures, global radiation and wind speed were data-logged at 20 second intervals. At the beginning of each test period the collector was pre-run for 15 minutes so as to reach a quasi-steady state. In analysing the data, only 5 minute periods when the average global radiation exceeded 800W/m2 and did not

1K and the inlet temperature did not vary by more than 0.1K were taken to be steady-state.

Additionally, any data points that satisfied these criteria but were more than 30 degrees either side of solar noon were also eliminated due to the possibility of including incident angle modifier terms.

Figure 20: Front panel of the labVIEW DAQ program.

2.4 Development of a BIT-SWH System

With the development of the BIT collector now complete, its applicability for SWH was investigated. There are many different SWH system configurations, the most common water heating systems is however well documented and reviewed elsewhere (Duffie and Beckman, 2006). In this study the SWH system that was chosen is based on the forced circulation (indirect or close looped, drain-back) system configuration. An overview of the system is shown if Figure 21. All the major components of the system shall be illustrated and discussed separately within this section (a schematic is provided in Appendix B).

Figure 21: A typical SWH forced circulation indirect (or closed loop), drain-back system (adapted from Harrison et al., 1985).

The difficulty with choosing an appropriate system configuration for this specific collector lies in the fact that current commercially available systems are packaged systems; which are optimised for use with traditional collectors (flat plate or evacuated tube). Since the design of the BIT collector closes matches that of a flat plate collector it was decided that the forced circulation configuration would be used. It was also chosen because it is a good representation of a typical SWH system in single family households in New Zealand (Weiss and Mauthner, 2012).

A study report by Pollard and Zhao (2008) surveyed a total of 39 SWH (and HPWH) systems in different locations within New Zealand. Of this, a total of 25 were pumped systems using either a flat plate or evacuated tube collector, as shown in Figure 22, thus supporting the use of this system configuration.

Figure 22: Surveyed houses by each technology defined by region (adapted from Pollard and Zhao, 2008)

As mentioned in the above section the collector used is unique, therefore the SWH system investigated in this study is unique. It utilises the BIT collector connected to a forced circulation drain-back configuration providing a novel BIT-SWH system. Another reason for choosing the forced circulation drain-back system is, pragmatically, it is a fail-safe method of ensuring that the collectors and collector loop piping never freeze. This is a major feature of the drain-back system,

the pump shuts off. Since the collector is already at a slight tilt to receive incident radiation, this allows the collector to be completely drained. A sight glass attached to the drain-back tank shows when the reservoir tank is full and the collector has been drained.