A data driven study on LEED credits and green building related implications in city planning

Leadership in Energy and Environmental Design (LEED) from the United States is one of the most popular green building standards around the world. The growing concern on global warming has led to increasing attention on LEED and green buildings in both industry and academia. Many studies have been conducted to analyze green building technologies and LEED project management, but the majority of the existing literature fails to address the following four major gaps and limitations: (1) LEED certifies green buildings into different grades according to the number of credit points each building has achieved. LEED managers often attempt to achieve as many credits as possible with limited budgets and resources. However, there is a lack of studies to explore the historical credit achievements and the relationship between credits in order to better facilitate the LEED project design. (2) The majority of related studies focus on the achievement of the credits in one specific green aspect, and there is a lack of research on studying project management or target credit selection from the whole picture, including all the other green aspects such as sustainable sites and water efficiency. (3) To better promote the development of green buildings, efforts from individual green buildings are not enough. Experts from the domain of city planning have been arguing that a top-down approach from the government to stimulate the green building markets would be a more efficient way. However, limited studies have been conducted to explore city planning from the perspective of LEED and green buildings. (4) The majority of the existing studies on LEED and green buildings focused on only one or two specific cases. This made their findings less general and may not be applicable to other green building projects. As a result, data driven studies on large numbers of green building projects are needed in order to discover more general patterns and implications. This research therefore aimed to study the credit achievements of individual green buildings (bottom-up) and related implications on green city planning (top-down) using data mining approaches. From the bottom-up aspect, in order to provide LEED managers a better understanding on the LEED credits so as to better achieve the target credits, the relationships between the credits were analyzed based on 1,000 certified LEED projects using the association rule analysis. Results show that some credits like SSc6. 2 (Stormwater Design) and WEc1 (Water Efficient Landscaping) are strongly related and can be achieved through single green building technology with little additional effort. In addition, in order to provide clearer examples with details about the target credits and green building technologies, a decision support web tool was developed based on a non-linear case-based reasoning framework to provide historical similar examples. Test results on 20 recently certified LEED projects showed that the system was able to accurately predict 80. 75% of the target LEED credits based on the retrieved similar projects. From the top-down aspect, since the location factor was discovered to be an important factor in identifying similar projects, it is suspected that the climate factors may affect the green building project design. Therefore, a study on identifying the influential climate factors on the LEED credits was conducted based on the Random Forests classification algorithm. Results provide useful suggestions to the government on mitigating the heat island effect. The climate factors were further extended to explore other features such as social and economic features on the development of LEED markets. The leading regional green building markets in the U.S. were identified using clustering analysis, and the influential features were investigated using classification analysis. Based on the results, it could be suggested to the Chinese government that 20 Chinese cities be the focus of development in regard to the perspective of learning from the U.S. experience. In addition, considering the energy asp ct is the most important sector in LEED according the weights of the credits, a case study based on 3,640 residential buildings in New York City was conducted to analyze the influence of 171 different features on the regional energy use intensity (EUI). Results reveal that the percentage of people with higher education is the most important feature affecting the regional EUI of residential buildings.