a b s t r a c tDirective 2002/91/EC of the European Parliament and Council on the Energy Performance of Buildings has led to major developments in energy policies followed by the EU Member States. The national energy performance targets for the built environment are mostly rooted in the Building Regulations that are shaped by this Directive. Article 3 of this Directive requires a methodology to calculate energy performance of buildings under standardised operating conditions. Overwhelming evidence suggests that actual energy performance is often significantly higher than this standardised and theoretical performance. The risk is national energy saving targets may not be achieved in practice. The UK evidence for the education and office sectors is presented in this paper. A measurement and verification plan is proposed to compare actual energy performance of a building with its theoretical performance using calibrated thermal modelling. Consequently, the intended vs. actual energy performance can be established under identical operating conditions. This can help identify the shortcomings of construction process and building procurement. Once energy performance gap is determined with reasonable accuracy and root causes identified, effective measures could be adopted to remedy or offset this gap.

This study presents a method for assessing energy efficient refurbishment options for schools in the UK. The method accounts for life cycle effects on cost and carbon emissions since refurbished buildings will last for many years.Four schools are identified as representative of school archetypes built in the UK during four distinct periods in the 20th century. The schools are used as a base for simulation of the effects of energy efficient refurbishment of building fabric and heating plant. All possible combinations of the selected measures are simulated. Simulated energy savings are then compared between the four school buildings, demonstrating how physical characteristics of the schools affect the available savings. Simulating combinations of energy efficiency measures allow analysis of interaction effects between measures, and reveals some positive and some negative interactions. A regression model of energy savings in the four schools is also developed.Simulated energy savings are then used as inputs for a life cycle assessment model. Life cycle indicators considered are marginal life cycle cost and marginal life cycle carbon footprint. These metrics are used to rank the energy efficiency measures on net present value and life cycle carbon footprint saving, both individually and in combination with each other.Carbon payback is shorter than financial payback in all scenarios, and all measures and combinations of measures repaid the carbon invested in them. Positive net present value is less common, and frequently depends on air tightness improvements also being achieved.

The bottom-up methods for energy benchmarking aim to derive a yardstick for energy performance based on a theoretical analysis of a building. While the top-down methods drive performance improvement by ranking a building against its peers, the bottom-up methods are focused on the building's specific context. Consequently, the bottom-up methods can help identify how performance improvement could be materialised. These two complementary approaches can improve design practice and facilities' management. Two bottom-up methods that could be used for energy benchmarking have been reviewed using UK schools as case studies: Building physics and aggregated end-use. The aim is to demonstrate how these methods could be used for benchmarking and identify their benefits and limitations. When all energy components are included in a model under expected operating conditions, the building physics method can be used to establish a baseline for energy performance. It is demonstrated that where this method is used under standardised operating conditions and is subject to minimum energy performance requirements, as prescribed by the Energy Performance of Buildings Directive (EPBD), it can be used to establish a benchmark for energy performance. It is also shown how aggregated end-use methods such as CIBSE TM22 can be used to define system level benchmarks, and identify the root causes for discrepancy between measured performance and design intent in a systematic way.

Studies have shown that the actual energy consumption of buildings once built and in operation is often far greater than the energy consumption predictions made during design-leading to the term: 'performance gap'. An alternative to traditional simulation methods is an approach based on real-world data, where bahaviour is learned through observation. Display Energy Certificates (DECs) are a source of observed building 'behaviour' in the UK, and machine learning, a subset of artificial intelligence, can predict global behaviour in complex systems, such as buildings. In view of this, artificial neural networks, a machine learning technique, were trained to predict thermal (gas) and electrical energy use of building designs based on a range of collected design and briefing parameters. As a demonstrative case, the research focused on school design in England. Mean absolute percentage errors of 22.9% and 22.5% for thermal and electricity energy use predictions respectively were achieved. This is an improvement of 9.1% for the prediction of thermal energy use and 24.5% for the prediction of electricity energy use when compared to sources evidencing the current performance gap. Fuel Type and Sector Mean Design Total Energy Consumption (kWh/m 2 /yr) Mean Actual Total Energy Consumption (kWh/m 2 /yr) Design Prediction Error (%): 'Performance Gap' Thermal Offices 46 73 37 Education 57 84 32 Electricity Offices 71 121 41 Education 56 106 47 Data-driven Approach "A major hindrance in modelling real problems is the lack of understanding of their underlying mechanisms because of complex and nonlinear interactions among various aspects of the problem [...] in many cases, the best solution is to learn system behaviour through observations" (Samarasinghe 2007, p.1-2). In view of this, an alternative approach at predicting energy consumption to mathematical models based on building physics (traditional building energy simulation) is to collect large amounts of actual energy and

Building performance evaluations (BPE) of five secondary schools and academies constructed under the Building Schools for the Future (BSF) programme in England found that CO 2 emissions associated with operational energy performance in all these buildings is higher than the median of the secondary schools. Whilst the new regulatory requirements for building fabric performance have led to some improvements in heating energy when compared against good practice and typical benchmarks, there is still significant discrepancy between heating energy use and the design expectations. Electricity use in these buildings is also 37-191% more than the median school and significantly worse than the design expectations. These results point to the importance of post-occupancy building fine-tuning and measurement and verification of performance in-use with respect to design projections to narrow the performance gap. It is also necessary to set out clear operational performance targets and protect energy efficiency measures from value engineering throughout building procurement and in operation to achieve good level of performance. Finally, it is suggested to adopt a holistic view of energy, environmental quality, and educational performance to have a better understanding of schools' performance and potential conflicts between energy efficiency measures and indoor environmental quality (IEQ).