Darren Robinson

Adaptive actions on shading devices in response to local visual stimuli

Based on 6 years of continuous measurements, we have analysed in detail the occupancy, thermal and visual parameters influencing actions on shading devices in order to derive an accurate model for the prediction of their usage in office buildings. This article begins by presenting some of the key findings from these analyses. Informed by other developments in the literature, we go on to propose an approach for a comprehensive stochastic model for simulating blind usage. This model is based on a Markov process taking rigorously selected predictors (initial blind status, indoor and outdoor illuminance) as input variables to predict lowering and raising actions performed by occupants. A separate sub-model then predicts the chosen shaded fraction. An assessment of the predictive accuracy of simulations is then presented for several modelling variants using our measured data, from which the best performing model variant is selected.

Verification of stochastic models of window opening behaviour for residential buildings

Based on the analyses of data from two distinct measurement campaigns conducted in residential indoor environments in Japan and Switzerland, we identify the specificities of occupants’ behaviour with respect to their interactions with windows, including the choice of opening angles for axial openings. As a first step, each dataset is analysed to develop separate predictive models which account for the specificities of window usage in the residential context. The predictive accuracy of these models is then challenged by validation on external data: using models inferred from data obtained from one survey, actions on windows are simulated for the other survey and the predictions are compared with observations. Dynamic models developed using data from office buildings as well as previously published models are also compared using this verification procedure. In the case of the Swiss dataset, these analyses demonstrate the ability of carefully formulated behavioural models developed from office environment data to reliably predict window usage in a residential context and vice-versa. However, we observe that the same models perform less satisfactorily in the prediction of window usage in Japan. From these results it seems that such models require specific calibration in the case of buildings equipped with an air-conditioning unit as was the case for the hot and humid summer climate of Japan.

Predicting the urban solar fraction: a methodology for energy advisers and planners based on GIS

This paper describes the development of the underlying methodology of a solar energy planning (SEP) system for energy advisers and policy makers. The methodology predicts the baseline energy consumption of domestic properties and determines the potential for reducing this using the three key solar technologies of passive solar design, solar water heating and photovoltaic (PV) systems. A new dwelling classification system has been developed to address the major problem of data collection for city-wide domestic energy modelling. The system permits baseline energy demands to be estimated using assumed values or more accurately calculated using dwelling survey data. The methodology integrates existing models with new approaches to both identify suitable dwellings for installing solar water heating and PV systems and to quantify the potential energy savings and reductions in carbon dioxide emissions. Guidance on improving estate layouts to enhance passive solar conditions is also given. Results can be presented using a geographical information system (GIS). The paper concludes with a discussion of possible planning scenarios to illustrate how the methodology may enable planners to consider the urban-scale application of solar energy with greatly increased confidence.

A hybrid CMA-ES and HDE optimisation algorithm with application to solar energy potential

This paper describes the results of initial experiments to apply computational algorithms to explore a large parameter space containing many variables in the search for an optimal solution for the sustainable design of an urban development using a potentially complicated fitness function. This initial work concentrates on varying the placement of buildings to optimise solar irradiation availability. For this we propose a hybrid of the covariance matrix adaptation evolution strategy (CMA-ES) and hybrid differential evolution (HDE) algorithms coupled with an efficient backwards ray tracing technique. In this paper we concentrate on the formulation of the new hybrid algorithm and its testing using standard benchmarks as well as a solar optimisation problem. The new algorithm outperforms both the standalone CMA-ES and HDE algorithms in benchmark tests and an alternative multi-objective optimisation tool in the case of the solar optimisation problem.

A simplified radiosity algorithm for general urban radiation exchange

The radiant external environment may be described by two hemispheres, above and below the horizontal plane, which are discretized into a number of patches of known solid angle. Occlusions to these patches may be combined and represented as some patch fraction for which the radiant characteristics are defined by the dominant occlusion. By solving for radiant exchanges between each surface in a scene and its associated (un)occluded patches, we have a simplified radiosity algorithm (SRA). This paper describes the application of this SRA to solve for urban scale predictions of: (i) solar radiation; (ii) interior daylight; and (iii) longwave radiation. Comparisons of solar and daylight predictions with the ray-tracing program RADIANCE show that accurate results are achieved at a computational cost several orders of magnitude lower. Practical application: This paper describes a new model for predicting external irradiance (shortwave and longwave) and internal illuminance in an accurate and very efficient way, in a single computational module. This module may be incorporated into existing software to improve the quality of predictions from single building thermal simulations as well as emerging software for urban scale predictions of integrated resource (energy, water, waste) flows, for which the model was developed.

A comparison of global optimization algorithms with standard benchmark functions and real-world applications using EnergyPlus

There is an increasing interest in the use of computer algorithms to identify combinations of parameters which optimise the energy performance of buildings. For such problems, the objective function can be multi-modal and needs to be approximated numerically using building energy simulation programs. As these programs contain iterative solution algorithms, they introduce discontinuities in the numerical approximation to the objective function. Metaheuristics often work well for such problems, but their convergence to a global optimum cannot be established formally. Moreover, different algorithms tend to be suited to particular classes of optimization problems. To shed light on this issue we compared the performance of two metaheuristics, the hybrid CMA-ES/HDE and the hybrid PSO/HJ, in minimizing standard benchmark functions and real-world building energy optimization problems of varying complexity. From this we find that the CMA-ES/HDE performs well on more complex objective functions, but that the PSO/HJ more consistently identifies the global minimum for simpler objective functions. Both identified similar values in the objective functions arising from energy simulations, but with different combinations of model parameters. This may suggest that the objective function is multi-modal. The algorithms also correctly identified some non-intuitive parameter combinations that were caused by a simplified control sequence of the building energy system that does not represent actual practice, further reinforcing their utility.There is an increasing interest in the use of computer algorithms to identify combinations of parameters that optimize the energy performance of buildings. For such problems, the objective function can be multi-modal and needs to be approximated numerically using building energy simulation programs. As these programs contain iterative solution algorithms, they introduce discontinuities in the numerical approximation to the objective function. Metaheuristics often work well for such problems, but their convergence to a global optimum cannot be established formally. Moreover, different algorithms tend to be suited to particular classes of optimization problems.  To shed light on this issue, we compared the performance of two metaheuristics, the hybrid CMA-ES/HDE and the hybrid PSO/HJ, in minimising standard benchmark functions and real-world building energy optimization problems of varying complexity. From this, we find that the CMA-ES/HDE performs well on more complex objective functions, but that the PSO/HJ more consistently identifies the global minimum for simpler objective functions. Both identified similar values in the objective functions arising from energy simulations, but with different combinations of model parameters. This may suggest that the objective function is multi-modal. The algorithms also correctly identified some non-intuitive parameter combinations that were caused by a simplified control sequence of the building energy system that does not represent actual practice, further reinforcing their utility.

Modelling occupants' personal characteristics for thermal comfort prediction

Based on results from a field survey campaign conducted in Switzerand, we show that occupants’ variations in clothing choices, which are relatively unconstrained, are best described by the daily mean outdoor temperature and that major clothing adjustments occur rarely during the day. We then develop an ordinal logistic model of the probability distribution of discretised clothing levels, which results in a concise and informative expression of occupants’ clothing choices. Results from both cross-validation and independent verification suggest that this model formulation may be used with confidence. Furthermore, the form of the model is readily generalisable, given the requisite calibration data, to environments where dress codes are more specific. We also observe that, for these building occupants, the prevailing metabolic activity levels are mostly constant for the whole range of surveyed environmental conditions, as their activities are relatively constrained by the tasks in hand. Occupants may compensate for this constraint, however, through the consumption of cold and hot drinks, with corresponding impacts on metabolic heat production. Indeed, cold drink consumption was found to be highly correlated with indoor thermal conditions, whilst hot drink consumption is best described by a seasonal variable. These variables can be used for predictive purposes using binary logistic models.

Passive Downdraught Evaporative Cooling I. Concept and Precedents

This is the first of a series of four papers that describe a 3-year EU-funded research project into the application of passive downdraught evaporative cooling to non- domestic buildings. In this paper various evaporative cooling techniques are reviewed. By spraying fine drop lets of water at the top of atria, a downdraught of air cooled by evaporation can be produced. Such direct eva porative cooling using an evaporation tower appears to be a suitable approach for partly displacing the need for air-conditioning in hot, dry climates. It can satisfy fresh air requirements and reduce or eliminate demand for mechanical cooling. Examples of this cooling technique in Southern Europe and the Middle East have already demonstrated its operation and potential energy sav ings. However, limitations, primarily due to control of the system, have been identified. This introductory paper presents the theoretical basis of evaporative cooling, reviews some historical precedents, and discusses their relative strengths and weaknesses. Three further papers in this series will disseminate the main findings of the project.

Modelling Occupants' Presence and Behaviour – Part I

The first generation of dynamic simulation programs (DSPs) was developed during the late 1970s and early 1980s. These were essentially command-line interfaces to routines to calculate the dynamic heat transfers within a building and between this and the outside environment. Subsequent work concentrated on improving the usability of these routines and extending the scope of the programs’ core capabilities, for example, to incorporate coupled plant and mass flow modelling. With improved functionality and amidst growing demand for their use by the more pioneering design consultants, attention shifted to proving the validity of their core heat transfer models. By the mid 1990s, with results from these validation studies taken on board and with improved usability, attention then focused upon the addition of further modelling functionality. This included the addition of 3D conduction modelling, links with ray tracing programs for improved lighting modelling, electrical power flow modelling and embedded computational fluid dynamics (CFD). The result is programs such as ESP-r (Clarke 2001) which has a transient finite difference heat flow solver at its core and supports simultaneous solutions of plant, fluid, electrical power and CFD equation sets. At the dawn of the new Millennium then, designers and academics alike had at their disposal highly sophisticated programs for simulating buildings’ energy and environmental performance. But these programs remained limited in that the presence and behaviour of building occupants were represented as simple deterministic profiles and rules. In this two part special edition of the Journal of Building Performance Simulation (JBPS), we confront this limitation, focussing on advances that have been made during the past decade to improve our understanding of the impact that occupants have on buildings’ performance and how this may be usefully encapsulated in algorithmic form. But before we embark on this venture, we should perhaps ask ourselves the simple question ‘do we really need to represent occupants in a more realistic way in our analyses?’ The answer, as ever, is that it depends on what we want to learn. If our purpose is simply to compare the relative energy performance of alternative design proposals, all other factors resting constant, then the answer may be – no, we do not. But if our purpose is to understand whether a particular design proposal is a robust one, or to determine whether this proposal is likely to be comfortable or to understand what its energy and environmental performance is likely to be in absolute terms, then the answer is most assuredly – yes, we do (an assertion that will be underlined on numerous occasions throughout the first part of this special edition). Furthermore, as the quality of buildings’ envelopes improves, thus better conserving energy, so the consequences of our occupants’ actions, to regulate the envelope and the internal lights and appliances accommodated, are further exaggerated. In short, it will be more and more important to accurately model occupants’ presence and behaviour as our buildings’ performance improves, particularly as we strive towards fully passive buildings with no dedicated heating or cooling system. This is our motivation behind this two part special edition. In this first part then, we place a particular focus on occupants’ impacts on buildings’ energy and environmental performance, using progressively more sophisticated and/or comprehensive representations of their presence and associated behaviour. To this end, Roetzel et al. open this special edition by examining, using EnergyPlus, the impact of occupants’ presence and behaviour on the energy and environmental performance of three variants of a mixed-mode office building located in Greece. In this they define pragmatic profiles and rules for the representation of occupants’ presence and their use of electrical appliances, blinds, lights and windows (this latter having a probabilistic component) according to both ideal and worst case scenarios. They conclude that both annual energy use and carbon dioxide emissions for the worst case scenario may exceed that of the ideal scenario by up to a factor of three and that indicators of indoor environmental quality may also vary significantly. Haldi and Robinson go a step further, integrating detailed stochastic models of occupants’ use of windows and shading devices into the urban energy modelling tool CitySim. Using this tool, they analyse, for different combinations of plausible changes in the building design and systems control parameters of Journal of Building Performance Simulation Vol. 4, No. 4, December 2011, 301–302

Passive Down-Draught Evaporative Cooling: Thermal Modelling of an Office Building

Two simulation-based methods of predicting the dynamic thermal performance of non-domestic buildings in hot dry climates, conditioned by Passive Downdraught Evaporative Cooling (PDEC), have been evaluated. In the PDEC system analysed, microscopic droplets of water were evaporated in ambient air and the resulting cool dense air delivered passively, through controlled openings, to occupied areas. The first method predicts the annual hours of PDEC operation, the water usage and the frequency of occurrence of different internal temperatures. The second method, which requires a thermal model with an integrated airflow network, provides insight into the airflow control strategies. Both methods demonstrated that an office building, in Seville, would need support from a mechanical cooling system to remain comfortable. However, the predicted energy use and carbon dioxide production was just 25% of that for a conventional airconditioned building.