Ireland’s housing stock is amongst the poorest performing in Northern Europe [1], therefore tackling energy efficiency measures in the domestic sector is of extreme priority. Each year, an estimated 30,000 to 60,000 excess deaths occur in the UK, while in Ireland 1,500 to 2,000 excess deaths occur due to winter mortality caused by cold strain [2]. With 44% of the Irish housing stock built prior to any energy specific Building Regulation [3], it is no surprise that problems associated with mould and damp are widespread. Subsequently, the retrofitting of the existing Irish Local Authority Housing (LAH) built stock would reduce unnecessary heat loss and improve exposed surface temperatures which eliminates or reduces the risk of condensation or mould growth. A study carried out by Fisk et al [4] found that there was a 30% to 50% increase in respiratory and asthma related health outcomes in buildings with dampness and visible mould problems. Further, if a major reduction in energy consumption in the domestic sector is to be achieved, various researchers have suggested that this reduction will have to come from Energy Efficient Retrofit (EER) measures applied to the existing housing stock [5] [6] [7] [8]. In order to rank and prioritise potential high impact EER solutions, there must first be a detailed understanding of the existing housing stock thermophysical characteristics. This knowledge can also assist in resolving issues which contribute to problems such as visible mould, caused by unacceptably low surface temperatures. Currently, no detailed database is in existence in Ireland which describes the geometrical and thermal configuration of the existing housing stock. The research discussed in this article set out to develop a high-speed remote cataloguing and geometrical data extraction methodology to build such a database. Replacing traditional measurement and surveying methods with a remote surveying approach allows for the rapid extraction of façade information, reducing the time and disruption normally involved in carrying out these surveys.

Virtual survey method

To provide an initial base level of information a remote surveying process was developed, named the Virtual Survey Method (VSM) [9], allowing a systematic classification of the various house types in the existing housing stock. [caption id="attachment_34509" align="alignright" width="300"]Fig 1a CLICK TO ENLARGE Fig 1a) Map of Cork Cities 36 LAH developments [10][/caption]Cork City’s LAH stock was used as a case study to test the VSM application, although its potential for wider application is effectively global. The VSM combines user features of Google Earth, Google Street View and OSI online maps to catalogue and disaggregate the housing stock. 75 LAH developments were initially identified through collaboration with the Housing Department at Cork City Council. A VSM application criterion was introduced using a vintage category from 1930 to 1982, as it was found that homes built from 1941 to 1979 have the highest number of fuel-poverty sufferers, affecting over 111,000 Irish homes. [caption id="attachment_34511" align="alignright" width="256"]Fig 1b CLICK TO ENLARGE 1b): 18 terraced house typologies identified across the 36 LAH developments[/caption] The housing stock also had to be of terraced formation, as those experiencing fuel poverty are more likely to live in semi-detached or terraced house-types and are highest amongst tenants than home owners. Having applied this criterion, Cork City LAH housing stock was refined to 36 LAH developments. Following a remote visual survey of the housing stock using Google Street View, some 18 different terraced-house typologies were identified, amounting to 10,487 LAH built units. Figure 1 highlights the 36 Cork City developments used in the remote surveying process, along with the 18 house typologies defined following the VSM. Each LAH development was analysed remotely and disaggregated into a number of terraced typologies: number of units within each typology, end of terrace, mid terrace, semi-detached, orientation and elevation. Further details of this work can be found in [9, 10].

Remote cataloguing, measurement and mapping method

[caption id="attachment_34519" align="alignright" width="300"]2a and 2b CLICK TO ENLARGE Fig 2a) GE measurement extraction from full terrace length and b) GE measurement extraction from single unit[/caption] Building on the information platform produced by the VSM, a remote cataloguing, measurement and mapping method (RCMM) [10] was developed to remotely extract detailed geometrical information from each terrace typology. Using a remote feature-extraction methodology to obtain geometrical information, this study has produced a detailed building component database premised on stock aggregation theory [11]. The RCMM uses photogrammetric techniques for measurement extraction, through the use of static panoramic viewports from Google Street View (GSV) and aerial spot images from Google Earth (GE). This approach speeds up the survey process (shown to be up to 75% faster than traditional approaches [8]) and generates a classification of the housing stock, and geometrical database remotely. In effect, the surveyor never leaves their desk. [caption id="attachment_34523" align="alignright" width="300"]2c CLICK TO ENLARGE 2c) Application of constraints to GSV image for geometry extraction[/caption] As building façades are premised on a Cartesian co-ordinate system [12], with each structure being orthogonal with one another, this offers itself to linear application of line primitive constraints [12]. The RCMM process is completed by utilising GE to extract roof measurements, using the GE distance finder, and GSV to obtain elevation information through a photogrammetric method. The operator must move the GSV camera position to line up with parallel and perpendicular line primitives, which are generated using an in-built software application (‘Rulers, Guides, Eye Dropper and Colour Picker 1.1.). [caption id="attachment_34525" align="alignright" width="300"]2d CLICK TO ENLARGE 2d) 2D drawing and 3D model output[/caption] Once symmetry has been achieved and the GSV image has been orthographically-rectified, the image can then be locked in place using a second software application (‘Google Page Ruler 2.0.9’). Google Page Ruler locks the image in place and allows the operator to extract measurements from the image in pixel format, with geometric building lines either parallel or perpendicular to the applied line primitives. Geometrical information for each house type can then be extracted from GSV images and processed in the same way to generate each façade model. The number of elevation images necessary is dependent on the length of the terrace and obstacles such as trees or cars. Figure 2 outlines an example of the RCMM in application. The RCMM has been shown to be sufficiently accurate for surveying purposes when compared to other methods.

RCMM, mapping and inventory system

[caption id="attachment_34527" align="alignright" width="300"]Fig 3 CLICK TO ENLARGE Fig 3: Referencing system (a) global plan orientation with terrace plan orientation example, (b) plan view of stepped terrace formation, (c) elevation view of stepped terrace formation built on slope, (d) elevation view disaggregation of 1, 2 and 3 stories formation applied to sample house types[/caption] A mapping and inventory system was also developed as part of this study to catalogue and quantify the disaggregation of the building envelope structure. A visual representation of the mapping of facades is outlined in Figure 3. The mapping method allows low-level cataloguing of individual facades. It is a basic building-stock database component identification system that deconstructs the various façade geometries into retrievable numerical codes within a database (i.e. for example using a relational database search query). Referenced areas are used to locate terrace façade sections under evaluation. It also generates a referencing platform for systematic exercises such as materials costing, pre-works design analyses and EER decision support. Referenced geometrical data from each terraced house typology can be identified, extracted and multiplied by totalised housing-unit values to support insights into the aggregated housing stock. The extracted information can then be used to calculate and quantify totalised values for an entire housing development. The system specifically addresses the cataloguing of construction details that contribute to problems such as thermal bridging, etc. Detailed information on linear measurements associated with typical thermal bridges is necessary for calculating the extra heat loss through typical problematic construction details. This information, in conjunction to construction detail quantity and location, make this a powerful tool for generating specific geometrical data and component area quantification for large-scale thermal modelling supporting high-level investment decisions across multiple housing developments. The objective is to know precisely the position and quantity of wall detail, window and door location and type of detailed geometry of the terraced-house envelope under evaluation. Again, this is achieved remotely, so can be integrated with other decision support tools such as energy and financial modelling packages. Further details are available in [10]. The combined application of the VSM, RCMM and statistical modelling using 1551 BER data files from the Sustainable Energy Authority of Ireland’s national BER research database [13] allowed the further distillation of all information related to the 10,487 houses into five representative house archetypes. These were then used for stock aggregated modelling of energy and CO2 performance of an externally applied high-performance thermal envelope retrofit upgrade. Results from the modelling are of the five archetypes and further detailed analysis is available from [9, 10].

Discussion and conclusion

This study has produced a detailed disaggregated building component database for a total of 10,487 LAH units in Cork City represented by 18 terraced-house typologies and, with further data from the BER research portal, five representative archetypes. A detailed component identification system has also been designed to map each terrace-house façade. The RCMM is suitable for integrating into any future automated process streamlining the surveying and information collection stages of database generation of a national building stock geometry. By embedding the mapping and inventory parts into the global framework it can be used as a basis for a building stock database retrieval system (coupled with photo and model libraries). The method is adaptable in structure and can be applied to a range of house type variations. Data systems and detailed analysis need to be embedded into national housing stock databases. A higher quality of extracted low level building data can lead to increased accuracy in outputs from housing stock evaluation studies, giving more realistic results from modelling large scale urban level EER measures thus contributing to a reduction in the energy performance gap. Authors: Primary author: Dr James Pittam Secondary author: Mr Paul D O’Sullivan Dr James Pittam currently works as a researcher at the zero 2020 testbed in CIT (nZEB).  He completed his PhD research at MeSSO developing remote surveying, cataloguing and measurement techniques for the disaggregation of the local authority housing stock into representative archetypes to facilitate large scale thermal modelling. Prior to CIT, Dr Pittam worked in industry as a passive house consultant for WainMorehead Architects, while carrying out an MSc in passive-house retrofit and is certified as a passive-house designer. He has an excellent understanding of the structural composition of existing buildings due to extensive experience in the construction industry. References 1. Lapillonne, B., Sebi, C, and Pollier, K, Energy Efficiency Trends for Households in the EU. Enerdata—An Analysis Based on the ODYSSEE Database, 2012. 2. Group, T.E., Cold exposure and winter mortality from ischaemic heart disease, cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. The Lancet, 1997. 349(9062): p. 1341-1346. Doi:10.1016/s0140-6736(96)12338-2 3. SEAI, Energy in the residential sector 2013 Report. Available at: http://www.seai.ie/Publications/Statistics_Publications/Energy-in-the-Residential-Sector/Energy-in-the-Residential-Sector-2013.pdf. 4. Fisk, W.J., Lei, Q., Gomez, and Mendell M.J. Meta-analyses of the associations of respiratory health effects with dampness and mold in homes. Indoor air, 2007. 17(4): p. 284-296. Doi:10.1111/j.1600-0668.2007.00475.x 5. Gupta, R., Moving towards low-carbon buildings and cities: experiences from Oxford, UK. International Journal of Low-Carbon Technologies, 2009. 4(3): p. 159-168. Doi: 10.1093/ijlct/ctp028 6. Bell, M. and Lowe, R, Energy efficient modernisation of housing: a UK case study. Energy and Buildings, 2000. 32(3): p. 267-280. Doi:10.1016/s0378-7788(00)00053-0 7. Beddington, J., Managing energy in the built environment: Rethinking the system. Energy Policy, 2008. 36(12): p. 4299-4300. Doi:10.1016/j.enpol.2008.08.028 8. Ahern, C., P. Griffiths, and Flaherty, M.O, State of the Irish housing stock—Modelling the heat losses of Ireland's existing detached rural housing stock & estimating the benefit of thermal retrofit measures on this stock. Energy Policy, 2013. 55: p. 139-151. Doi:10.1016/j.enpol.2012.11.039 9. James Pittam, Paul D. O'Sullivan, Garrett O'Sullivan, Stock Aggregation Model and Virtual Archetype for Large Scale Retro-fit Modelling of Local Authority Housing in Ireland, Energy Procedia, Volume 62, 2014, Pages 704-713, ISSN 1876-6102, http://dx.doi.org/10.1016/j.egypro.2014.12.434. 10. James Pittam, Paul D. O'Sullivan, Garrett O'Sullivan, A remote measurement and mapping technique for orderly rapid aggregation of building stock geometry, Automation in Construction, Volume 71, Part 2, November 2016, Pages 382-397, ISSN 0926-5805, http://dx.doi.org/10.1016/j.autcon.2016.08.013. 11. Moffatt, S., Stock Aggregation: Methods for Evaluating the Environmental Performance of Building Stocks. Annex 31, 2004. Energy-Related Environmental Impact of Buildings. Available at: http://www.iisbe.org/annex31/bkgrnd_reports.htm 12. Zhong, B., Xu, D. and Yang, J. Vertical corner line detection on buildings in quasi-manhattan world. in Image Processing (ICIP), 2013 20th IEEE International Conference on. 2013. IEEE. Doi:10.1109/icip.2013.6738631 13. SEAI, National BER Research Tool. Available from: http://www.seai.ie/Your_Buil ding /BER/National_BER_Research_Tool/ (9/02/16).