Optimization of significant insolation distribution parameters e A new approach towards BIPV system design D. Paul a, * , S.N. Mandal b , D. Mukherjee c , S.R. Bhadra Chaudhuri d a SSBB & Senior Member-ASQ, Kolkata, India b Kalyani Govt Engg College, Kalyani, India c Dept of E. & T. C. Engg, B.E.S.U., Shibpur, India d Dept of E. & T. C. Engg, B.E.S.U., Shibpur, India article info Article history: Received 16 December 2008 Accepted 25 February 2010 Available online 26 March 2010 Keywords: Solar radiation Building integrated photovoltaics Statistic Frequency distribution Quality function deployment Artificial neural network abstract System efficiency and payback time are yet to attain a commercially viable level for solar photovoltaic energy projects. Despite huge development in prediction of solar radiation data, there is a gap in extraction of pertinent information from such data. Hence the available data cannot be effectively utilized for engineering application. This is acting as a barrier for the emerging technology. For making accurate engineering and financial calculations regarding any solar energy project, it is crucial to identify and optimize the most significant statistic(s) representing insolation availability by the Photovoltaic setup at the installation site. Quality Function Deployment (QFD) technique has been applied for iden- tifying the statistic(s), which are of high significance from a project designer's point of view. A MATLABÔ program has been used to build the annual frequency distribution of hourly insolation over any module plane at a given location. Descriptive statistical analysis of such distributions is done through MINITAB TM . For Building Integrated Photo Voltaic (BIPV) installation, similar statistical analysis has been carried out for the composite frequency distribution, which is formed by weighted summation of insolation distributions for different module planes used in the installation. Vital most influential statistic(s) of the composite distribution have been optimized through Artificial Neural Network computation. This approach is expected to open up a new horizon in BIPV system design. Ó 2010 Elsevier Ltd. All rights reserved. 1. Introduction The most significant ingress of renewable energy into the modern urban life has just been initiated through the installation of BIPV systems, which is essentially deployment and integration of large area PV panels with all possible facades of buildings [1]. Accordingly, PV Array configurations should be ‘space-intelligent’ [2] as well as ‘insolation-intelligent’ in nature in order to ensure true energy sustainable and economically viable [3] design. But like most other application areas of renewable energy technologies, some engineering gaps are still prevalent resulting in uncertainty of performance of BIPV energy systems. In fact this kind of techno- management shortcoming is primarily responsible for the inability of solar PV technology to attract adequate investment [4] and business intelligence required for its desired market penetration. This critical gap between laboratory and life for an emerging technology can be effectively bridged by Quality Engineering tools and techniques, which are in many cases based on proven statistical methods. Using such statistical methods, an extremely use friendly and economic yet accurate instrument for insolation measurement have been successfully developed [5]. In another application, an algorithm based on Quality Engineering tools like Designed Experiment (DOE), ANOVA, Regression modeling and Response Surface Methodology (RSM) was developed for accurate prediction of PV module behavior under any given environmental condition [6,7]. Further research showed that the term “given environmental condition” calls for objective quantification of the insolation avail- ability for a site and to be more precise, for the solar photovoltaic installation at that site. It is to be noted that the ‘insolation avail- ability at a site’ and ‘insolation availability by a Solar PV installation’ are not synonymous. Discussion in the subsequent sections will show why lack of clarity in understanding and estimating the most appropriate ‘Statistic(s)’ representing the insolation availability and its variation may lead to wrong selection of PV modules, over specification, cost escalation and error in performance prediction of the solar PV energy systems. In this paper, a new way of defining and maximizing the ‘insolation availability’ on a solar PV installa- tion at a project site has been developed following three steps: * Corresponding author. E-mail addresses: prof_srbc@yahoo.com, debkalyan_paul@yahoo.co.in (D. Paul). Contents lists available at ScienceDirect Renewable Energy journal homepage: www.elsevier.com/locate/renene 0960-1481/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2010.02.026 Renewable Energy 35 (2010) 2182e2191