Organizational Unit:

School of Architecture

Permanent Link

https://hdl.handle.net/1853/70770

Parent Organization

Organizational Unit

College of Design

ArchiveSpace Name Record

https://finding-aids.library.gatech.edu/agents/corporate_entities/1120

Full item page

Publication Search Results

Now showing 1 - 10 of 27

Integrated assessment of buildings and distributed energy resources (DER) at the neighborhood scale

(Georgia Institute of Technology, 2017-11-13) Carneiro, Gustavo Antonio

In urban regions, traditionally a main electric grid fed by centralized power plants serves the growing energy demand of residential and commercial buildings. However, the advent of new technologies, such as distributed renewable energy generation, local energy storage, and smart controls, is transforming the way buildings interact and transact with the electric grid. When operating in coordination, several buildings or households can leverage their aggregate potential and use their energy flexibility and distributed resources to improve the operation of both the main grid and the pool of integrated and intelligent buildings. Much attention has been drawn to the potential benefits of these types of integration, especially the capabilities they can provide in terms of aggregate demand management and local power resilience. Nevertheless, building energy modeling at the urban level has not yet reached the necessary computational manageability and simulation robustness to assess these novel scenarios. To address this hiatus, the current thesis presents a computer-aided energy simulation method to model the integration of multiple buildings and distributed energy resources (DER) at the neighborhood scale. The proposed methodology uses a reduced order simulation approach to achieve a reliable and tractable dynamic modeling framework that can manage multiple transacting building energy models and DER models in a single platform. To test the modeling approach, this study first carries out a virtual experiment of a small community in Miami, FL, where it is possible to compare the outcomes of community energy consumption from our reduced order model to the outcomes from a higher order simulation approach. When using the community energy model to evaluate the performance of different DER options for community peak load shaving, we can observe that the influence of the model order reduction reveals to be very minor when compared to other uncertainties related to scenario variability and, especially, systems’ efficiencies. Secondly, we apply the reduced order modeling approach to an existing residential community in Rancho Cordova (Sacramento County), CA, with solar energy generation and battery energy storage. With this case study, we demonstrate the viability of our approach to construct and calibrate a reduced order model of fifteen households based only on limited and general data related to energy performance of the entire neighborhood. The developed reduced order model is used to evaluate the performance of different energy storage arrangements for reducing the occurrence of community super peak loads. In this virtual experiment, we can demonstrate how the model allows for uncertainty analyses over the influence of input parameters, as well as for more sophisticated optimization studies, including stochastic optimization, in a timely and transparent fashion. Finally, the proposed reduced order simulation approach is used to construct and test relevant energy performance measures at the neighborhood scale. Using the model unique features of manageability, reliability and flexibility, we propose the foundations for quantifying and measuring “community energy resilience” for outage situations, based on concepts of number of sustained hours and respective energy end-use convenience levels. We also measure and monetize DER options for providing “community energy flexibility”, aimed at shaping the load profile of a residential community to match the electric grid needs.
Addressing data informativeness in risk-conscious building performance simulation applications

(Georgia Institute of Technology, 2017-08-02) Li, Qi

Building performance management remains an important aspect in reducing building energy consumption and enhancing occupants’ thermal comfort and work productivity. Recent decades witnessed the maturity and proliferation of numerous methods, software and tools that span the whole spectrum of common building performance management practice. Among those related research and applications, the use of information and data in calibration and validation of building performance simulation (BPS) models constitutes an important subject of study especially in fault detection, operations management, and retrofit analysis. An extensive review of BPS model calibration and validation studies reveals two major research gaps. First, contemporary model calibration practice calls for an effective and robust method that can systematically incorporate a variety of information and data, handle modelling and prediction uncertainties, and maintain consistent model performance. Second, current approaches to collecting information and data in real practice largely depend on individual experience or common practice; further study is needed to understand the value of information and data, i.e. assess data informativeness, such as to support specific decision-making processes in choosing data monitoring strategies and to avoid missed opportunities or wasted resources. To this end, this dissertation develops a new framework to address data informativeness in model calibration and validation to answer two major research questions: 1) how to make optimal use of available information and data to calibrate a building simulation model under uncertainty, and 2) how to quantify the informativeness of information and data for risk-conscious building performance simulation applications. This framework builds upon uncertainty propagation using detailed measurements, and inverse modelling using Bayesian inference. It introduces probabilistic performance metrics to assess model prediction consistency and quantify data informativeness. Following an explanation of the framework’s theoretical soundness, this dissertation provides two case studies to demonstrate its practical effectiveness. The first is a controlled experiment in the Flexlab test facility at Lawrence Berkeley lab. A new validation methodology is proposed to validate a simulation model under uncertainty, in which the validation criteria build upon the introduced probabilistic performance metrics. Given the experiment setup, uncertainty propagation based on synthetic measurements is applied, which effectively improves prediction agreement and reduces the risk of accepting invalid simulation outcomes. The second is to determine the appropriate model form and metering data for a hypothetical intervention analysis of an existing building with hydronic heating on the Cambridge, UK campus. A three-level modelling method is proposed to enable modelling all thermal processes occurring in individual rooms while efficiently modelling the whole building to estimate heating system performance. Different sets of metering data are then used to calibrate the physical model, and the result indicates the superiority of Bayesian inference in exploiting the value of data, the necessity of room temperature and electricity monitoring under uncontrolled conditions, and the potential of daily metering data for calibration in real building performance management practice.
Optimal strategies for demand charge reduction by commercial building owners

(Georgia Institute of Technology, 2017-05-23) Zhang, Yuna

A substantial part of electricity bills in various types of commercial buildings, such as office buildings, hospitals and retails can consist of demand charges. Demand charges represent the penalty for an electricity consumer levied by the utility provider. They are typically a direct result of the shape of the power duration curve, in particular, the hours that a certain power level is exceeded in a given billing period (normally a month). Lowering the peak and/or reducing the hours that a power threshold is exceeded can drastically reduce demand charges. The ability to do so by dynamic, operational adjustments reflects the “energy flexibility” of the building. This term is now widely used in Europe and is the subject of a new international effort (IEA Annex 67) in this area. This thesis targets the optimal choice among design and operational measures in a retrofit or new design project that delivers the most effective way of reducing demand charges and increasing energy flexibility of commercial buildings. This goal will be achieved through an analysis of all feasible energy and peak reduction measures in different building types and in different use contexts. A search algorithm that compares all possible interventions will deliver the optimum, first with a deterministic analysis then with the recognition of the effects of all possible sources of uncertainty. This thesis evaluates the measures that are commonly adopted to decrease energy consumption and increase energy flexibility and thus reduce demand charges, including (1) upgrading building components and installing energy efficient equipment; (2) applying dynamic building load control strategies such as demand-side management; (3) installing a rooftop photovoltaic (PV) panel array. Operational interventions include the manipulation of thermostat settings and possibly the voltage reduction of lighting and appliances (in some cases including HVAC components) in the building, which may reduce thermal and visual comfort for certain periods. In order to support retrofit and design improvement decisions, an approach is developed that finds the optimal mix of measures that maximize the net present value of the investment in energy flexibility measures over twenty years for the owner. This study will analyze optimal solutions for three commercial building types. Differences between them in terms of energy use and peak demand will be investigated and a generically applicable measure of energy flexibility will be developed. These three buildings are chosen (by scaling their total floor area) such that their demand charges are in the same range. The monetary benefit of energy flexibility will be studied under different demand charge rate structures and under variable building consumption scenarios. This research will result in a new optimization framework for choosing the optimum among multiple options. Based on the proposed framework, this research will determine optimal ways to increase energy flexibility, leading to the best investment decisions for different commercial building types in different locations and under different rate structures.
Integrated performance based design of communities and distributed generation

(Georgia Institute of Technology, 2016-08-26) Street, Michael

The vertically integrated utility market within the U.S. is undergoing rapid changes due to the rise of small-scale distributed power generation known as microgrids, which are local networks of power generation and distribution typically serving a demand less than 40 MW. Primary drivers for microgrid investment are the performance benefits these systems return to their owners, which include increased reliability, reduced emissions and reduced operating costs. We define a novel modeling methodology to represent the microgrid as an integrated system of the demand and supply. Previous work to develop an integrated system model does not adequately model the building thermal demand, incorporate a modeler’s knowledge of the grid’s availability or allow for a user to model their tolerance for unmet demand. To address these modeling issues, we first demonstrate a technique for representing a building stock as a reduced order hourly demand model. Next, as demand side measures are typically defined at the building level as discrete options, we demonstrate a technique for converting a large discrete optimization problem into a simplified continuous variable optimization problem through the use of Pareto efficient cost functions. The reduced problem specification results in 90% fewer function evaluations for a benchmark optimization task. Then, we incorporate two new features into the Distributed Energy Resource Customer Adoption Model (DER-CAM) developed by Lawrence Berkeley National Laboratory (LBNL) that allow users to define grid outage scenarios and their limit of expected energy demand not served. Applying the integrated model to a microgrid design scenario return solutions that exhibit on average an 8% total annual cost reduction and 18% reduction in CO2 emissions versus a Supply Only case. Similarly, the results on average reduce total annual cost by 5% and annual emissions by 17% for a Demand First case. In summary, we present a modeling methodology with application to joint decision making that involve renewable power supply, building systems and passive building design measures and recommend this model for performance based microgrid design.
The impact of occupant modeling on energy outcomes of building energy simulation

(Georgia Institute of Technology, 2016-08-02) Kim, Ji Hyun

The reported performance gap between predicted and real building energy consumption has drawn keen attention from the building simulation community and related stakeholders. Alongside other research efforts to identify, quantify, and close this gap, the most recent attempt is the development of occupant behavior models that generate more “realistic” occupant inputs in the building energy simulation used for prediction. These new occupant models are typically realized by stochastic methods. To date, the newly developed models focus on mimicking real life variability. In spite of that, they have not led to more accurate consumption predictions than previous methods. Rather than adding yet another occupant behavior modeling approach, this thesis emphasizes the need to understand the impact of occupant models on building energy outcomes in real life applications. To accomplish this, we investigate two distinctive approaches to occupant modeling: top-down and bottom-up. We build the argument in the thesis that the top-down approach is suitable in highly variable situations where relatively little information about actual occupant variables can be known. This is usually the case in residential applications. By introducing a so-called “Life Style Factor,” we conclude that the use of this factor is promising to capture the variability of occupant-related parameters in residential buildings. For commercial buildings, a fundamental analysis is conducted to identify the impact of occupant-related inputs on the performance gap while explicitly considering the level of modelers’ knowledge about occupants’ presence and actions at the time of prediction. The results of a sensitivity analysis reveal that even in the case where the modelers’ ignorance of actual occupancy is significant and hence occupant parameters become important contributors to the performance gap, the resulting disparity could be fairly well quantified without introducing complex occupant behavior models. It is also found that the randomness of occupant behavior with respect to actions has no significant role in the performance gap at least in typical building simulation practice. This finding is significant as it advises us to rethink our pursuit of accuracy by developing new occupant behavior models, such as the ones that explicitly model the human reasoning, perception and action related to the opening of windows.
Accuracy, validity and relevance of probabilistic building energy models

(Georgia Institute of Technology, 2016-08-02) Wang, Qinpeng

Residential and commercial buildings consume 41% of total U.S. energy consumption. Since improving energy efficiency is still the most cost efficient energy saving option in the U.S., it is not surprising that many new buildings represent a push towards ultra-efficiency. Many studies argue that this calls for the use of high fidelity prediction models that by necessity will be probabilistic in nature due to many sources of uncertainty that affect the translation of a design specification into the actual reality of a constructed and operated facility. To inspect the fidelity of these probabilistic models against traditional deterministic models, we pose questions that address three major aspects of this new generation of building energy models: • Accuracy: do these models give more “correct” answers? • Validity: do these models lead to “better” design/retrofit decisions? • Relevance: does a profession that deploys these models provide “higher” value to the industry? This dissertation addresses the first question by identifying gaps in our understanding and quantifying various sources of model uncertainty reported in recent literature. Insufficiently understood and not well-quantified sources are further studied and resolved. The results of the above are analyzed in a sensitivity analysis that ranks input parameters alongside with model form uncertainties. Next, we adapt proven methods to conduct verification of probabilistic building energy models. Probabilistic calibration, marginal calibration and a continuous rank probability score are used to evaluate the “correctness” of the new generation of models. We illustrate the challenges of delivering validity proofs in a case study where outcomes of uncertainty analysis are translated into (monetary) risks and their influence is analyzed in a decision-making scenario involving energy performance contracts. Lastly, the study introduces a speculative approach to proving relevance by quantifying the overall societal benefit of a transparent risk framework that has the potential to unlock currently stagnating capital flow into large-scale building retrofits.
Application of inverse modeling to performance-based architectural design in the early stage

(Georgia Institute of Technology, 2016-05-31) Rezaee, Roya

The architecture, engineering, and construction community is taking action to reduce energy consumption. Fulfilling energy performance requirements entails complex decision-making at the architectural design stage, when a large number of parameters are undecided and the level of uncertainty is high. The early stage of design, in particular, is characterized by its iterative nature of divergent phases in which design alternatives are generated and convergent phases in which alternatives are assessed and selected. It is during or at the end of these phases that decision-making occurs under considerable uncertainty. Therefore, the methods and tools applied during these phases should account for the iterative, complex, and uncertain characteristics of the design process. At present, the building industry lacks a consistent approach to decision making during the phrases of the early stage of design: The divergent phase, when concepts are generated, consists of no practical framework within which designers generate more promising alternatives regarding energy performance, and the convergent phase, when concepts are evaluated and selected, includes no algorithm within it that designers can use to validate their decisions and provide confidence in their decisions. These deficiencies necessitate a clear step-wise approach that supports the proper design exploration by generation and evaluation of design alternatives, highlights significant parameters regarding energy performance for a variety of design scenarios, allows for coupled decisions under uncertainty, and align with the iterative nature of design process. This research hypothesizes that (1) a new systematic method based on linear inverse modeling (LIM) can generate plausible ranges for design parameters given a preferred thermal energy performance at the early stage of architectural design; and (2) the application of the proposed approach can lead to a higher probability of achieving energy efficient buildings (increase the chances of developing promising concepts), which is the main objective of performance-based design; and finally (3) in comparison to the current prescriptive approach, the proposed performance-based method help designers with the design process by providing more design freedom and guidance. Such an approach also accounts for the iterative nature of an architectural design and promotes a step-by-step procedure for making a decision and updating information as each new decision is made. In contrast to the conventional “forward modeling” in building performance analysis in which the design parameters are considered input and the energy performance are output, the “inverse modeling” deals with the performance objective as input and the design parameters are inferred as the output of the analysis. The study practices the proposed inverse modeling approach for making decisions regarding energy performance at the early design stages in four case studies, representing two different types of buildings in four climate zones. Such practices show the capability of the proposed inverse modeling to help designers in design space exploration, sequential decision-making, and trade-off study at the early stage of design. This method is proven to be a validate candidate for fulfilling desired energy performance and provide guidance and freedom in building design process. This thesis research contributes to the body of knowledge pertaining to building energy modeling and decision making at the early design stage, and its framework can be used by all groups of designers, the energy analysis experts as well as non-energy-expert architects, for a more informed decision-making regarding energy.
Performance measures for residential PV structural response to wind effects

(Georgia Institute of Technology, 2016-05-27) Goodman, Joseph Neal

This thesis applies structural reliability measures for the performance based design of residential PV system structures. These measures are intended to support designers in delivering systems with quantified and consistent reliability. Existing codified practices prescribe global factors (allowable stress design) and partial factors (load and resistance factor design) intended to provide an acceptable level of reliability as defined by historical practice. When applied to residential PV systems this prescriptive approach has two flaws, (1) calibration efforts needed to ensure consistency across structural system types have not kept up with the commercially available system types and (2) the actual expected reliability is not quantified and available to support decisions. The proposed reliability measures include probability of failure conditioned to wind speed in a fragility curve and the reliability index β, both of which are commonly used in performance based design. The approach is demonstrated through the application of the reliability measures to code compliant designs. Diverse system types are utilized to demonstrate how the existing code prescribed approach may lead to non-uniform structural performance. For each of the system types on which the reliability measures are demonstrated, a code compliant design is developed for three roof slopes, wind tunnel testing is conducted to provide an experimental measure of wind pressure coefficients, system specific fragility curves are generated to quantify the probability of failure conditioned to a set of wind speeds, and then, a site specific wind model is applied to produce a probability of failure and reliability index β. Through the performance based approach proposed in this thesis, two key outputs show non-uniform and unanticipated structural performance of PV systems designed according to the prescriptive code method. The two key outputs which illustrate this finding are fragility curves which illustrate the probability of failure over a range of wind speeds and reliability index, β values which couple the structural and wind distributions for a single measure of reliability.
Power performance assessment of building energy systems

(Georgia Institute of Technology, 2016-05-20) Makhmalbaf, Atefe

Buildings are the main consumers of electricity across the world. In the past research, the focus has been on evaluating the energy performance of buildings whereas the instantaneous power consumption of systems and aggregated load profiles have received less attention. Today, buildings are involved in the challenges of ‘power grid modernization.’ This is mostly because the increasing diversity of building systems requires a better understanding of their behavior during peak hours and the “demand charges” that are associated with it. Other drivers are the need to lower the carbon footprint of the electricity supply (i.e., reduction of grid as well as building scale emissions) and the growing number of demand response (DR) programs that rely on dynamic adjustments of building systems to support grid stability and resiliency. However, we lack methods, models, and performance measures that support building-grid interaction evaluations. This thesis has developed methods and models needed to study and assess performance of buildings in the electricity system. To achieve this, building thermal models, conventionally used to capture energy consumption are enhanced with electricity characteristics (e.g., voltage). With these models the impact of voltage on load shape of different systems is investigated and a set of quantitative power performance indicators (PIs) defined. These PIs are consequently applied to a variety of building control strategies in the context of DR scenarios. The developed PIs provide the fundamental component needed in decision support and auto-DR systems to quantitatively, systematically, and consistently compare and assess power performance of different building system types in given operation scenarios. This assessment is important for a range of applications. At building level, facility managers can use quantitative performance comparison of control strategies for both energy efficiency and peak reduction decisions. At grid level, our method can be used for power planning and management studies such as load forecasting. In the first part, this thesis demonstrates the feasibility of the thermal enhanced models with electrical characteristics by developing these models and showing how they can be constructed and used for different system types. In the second part, this thesis verifies usability of the performance assessment framework developed for DR and energy management decisions at building level. This is achieved by applying performance indicators defined to a set of scenarios. Results indicate how each performance indicator can support different performance criteria such as power and energy efficiency while maintaining thermal comfort of occupants. These quantitative PIs can be implemented in decision support systems that consider the trade-off between energy efficiency and investments in power management at the building site.
Closing the building energy performance gap by improving our predictions

(Georgia Institute of Technology, 2014-06-30) Sun, Yuming

Increasing studies imply that predicted energy performance of buildings significantly deviates from actual measured energy use. This so-called "performance gap" may undermine one's confidence in energy-efficient buildings, and thereby the role of building energy efficiency in the national carbon reduction plan. Closing the performance gap becomes a daunting challenge for the involved professions, stimulating them to reflect on how to investigate and better understand the size, origins, and extent of the gap. The energy performance gap underlines the lack of prediction capability of current building energy models. Specifically, existing predictions are predominantly deterministic, providing point estimation over the future quantity or event of interest. It, thus, largely ignores the error and noise inherent in an uncertain future of building energy consumption. To overcome this, the thesis turns to a thriving area in engineering statistics that focuses on computation-based uncertainty quantification. The work provides theories and models that enable probabilistic prediction over future energy consumption, forming the basis of risk assessment in decision-making. Uncertainties that affect the wide variety of interacting systems in buildings are organized into five scales (meteorology - urban - building - systems - occupants). At each level both model form and input parameter uncertainty are characterized with probability, involving statistical modeling and parameter distributional analysis. The quantification of uncertainty at different system scales is accomplished using the network of collaborators established through an NSF-funded research project. The bottom-up uncertainty quantification approach, which deals with meta uncertainty, is fundamental for generic application of uncertainty analysis across different types of buildings, under different urban climate conditions, and in different usage scenarios. Probabilistic predictions are evaluated by two criteria: coverage and sharpness. The goal of probabilistic prediction is to maximize the sharpness of the predictive distributions subject to the coverage of the realized values. The method is evaluated on a set of buildings on the Georgia Tech campus. The energy consumption of each building is monitored in most cases by a collection of hourly sub-metered consumption data. This research shows that a good match of probabilistic predictions and the real building energy consumption in operation is achievable. Results from the six case buildings show that using the best point estimations of the probabilistic predictions reduces the mean absolute error (MAE) from 44% to 15% and the root mean squared error (RMSE) from 49% to 18% in total annual cooling energy consumption. As for monthly cooling energy consumption, the MAE decreases from 44% to 21% and the RMSE decreases from 53% to 28%. More importantly, the entire probability distributions are statistically verified at annual level of building energy predictions. Based on uncertainty and sensitivity analysis applied to these buildings, the thesis concludes that the proposed method significantly reduces the magnitude and effectively infers the origins of the building energy performance gap.