Contribution by R. J. Hacker
The paper presents the results of detailed analysis of the energy performance and cost of energy produced for two complete photovoltaic (PV) systems installed as part of the fabric of the Zicer building at the University of East Anglia, UK. There are many published results of PV performance but it is often difficult to compare systems and understand the reasons for what are often very different performances. The authors have given a very clear exposition of the main factors which affect performance of these systems and extended the results to estimate performance in other UK locations.
The authors have noted the conflicting results which have been reported from previous embodied energy studies and produced their own results, again explaining some of the important variable parameters. The results, however, need to be understood in the context of the total world production of PV modules, of which these particular PV modules represent a small proportion. These modules are specially made for architectural integration with a semi-transparent aspect to allow light to enter the building and double-glazed construction to meet the Building Regulations (thermal and safety regulations). The large majority of modules manufactured have tightly packed cells and a single glass sheet with another material for backing and probably a metal frame for sealing the edges and mounting. There are also thin film PV cells available which are less energy intensive to manufacture. Consequently, it is possible that the results are not representative of the majority of modules which would be expected to have higher energy-yield ratios. It would be interesting to know if the authors have conducted an analysis of this factor.
Author's reply
The analysis focused solely on crystalline PVs due to the vast amount of actual monitoring data that was collected between 2003 and 2007 on the two crystalline arrays integrated into the Zicer building. The PV model that was developed for this work was based on the actual PV results from the monitoring programme and the model was not adapted to consider thin film technology.
However, altering the architectural design of the PV installation was considered as part of the research. In total the Zicer PV modules cover an area of 350 m2 but only 238 m2 are covered by the photovoltaic cells. The solar cells used on the faµade have a ‘packing factor’ of 62%, meaning that 38% of the PV module is open space between the cells, that is, clear glass. These large gaps between the PV cells were an architectural feature and usually for this type of installation the packing factor is selected on the basis of how much daylight is required to be transmitted through the modules. The PV roof has a slightly higher packing factor with 71%, a standard packing factor at the time in 2001.
Currently, standard packing factors for crystalline PV cells are between approximately 85% and 90%. If the PV cells on the faµade had packing factors of 85% then the average annual electricity generation from the faµade would increase from 2860 kWh to an estimated 4030 kWh (assuming no shading), a 41% rise in the polycrystalline PV electricity generation. Similarly, if the PV cells on the Zicer roof had packing factors of 85% then the annual average electricity generation would increase from 22 300 kWh to an estimated 26 910 kWh, a 21% rise in monocrystalline PV electricity generation. The increased PV generation from a higher packing factor would play a role in increasing the energy yield ratio and reducing the cost per Wh, although we have not considered the extent to how much the energy yield ratio would increase by/costs reduce by.
