Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

 
Session Overview
Session
B&I-3: Building Energy
Time:
Thursday, 27/Jun/2019:
10:30am - 12:00pm

Session Chair: Ehsan Vahidi
Location: Ross Island/Morrison

Presentations
10:30am - 10:50am

Assessing the impacts of energy code evolution on energy performance and embodied environmental impacts of multi-family residential buildings in the US

Ehsan Vahidi1, Randolph Kirchain2, Jeremy Gregory1

1Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139; 2Materials Research Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139

The buildings sector represents a significant fraction of greenhouse gas (GHG) emissions and energy consumption, making up 40% of the total global energy demand in both developed and developing countries. It is even larger when considering the entire footprint of the building sector, including embodied GHG emissions from raw material extraction, building material manufacturing, and on-site construction. Therefore, a life cycle perspective when studying the environmental impacts of buildings is imperative because any reduction in operational and embodied GHG emissions by the building sector can lead to significant environmental and economic gains.

Improvements to building energy efficiency through more stringent energy codes are often proposed as a mechanism for reducing the GHG emissions in the building sector. Building life cycle assessments (LCA) have shown that GHG emissions due to energy consumption represent the largest fraction of the life cycle for conventional building designs and thus, improving energy efficiency is a good strategy to lower GHG emissions. However, improving energy efficiency is often associated with more insulation and improved building materials and thus, it is expected that the embodied phase in the whole life cycle footprint is increasing.

The objectives of this analysis are to understand the potential for changes in building energy codes to reduce GHG emissions in the US and the impacts such changes will have on embodied emissions in building materials and construction. We use the case of multi-family residential buildings and conduct analyses for 14 cities representing each of the ASHRAE climate zones. Different design scenarios using both 2009 and 2015 codes developed by the International Energy Conservation Code (IECC) were investigated. Furthermore, to study future buildings that will need to comply with stricter codes, an energy efficient design scenario was also studied. Subsequently, considering the number of new multi-family residential buildings constructed in the United States, the impacts of building codes and standards on buildings’ embodied as well as operational GHG emissions were quantified in each state.

A 100,000 sqft multi-family residential apartment building with mixed one and two-bedroom units was analyzed in different locations over a life span of 50 years. Significant effort was invested to establish and account for a variety of design scenarios, including building envelope requirements for wood and insulated concrete form (ICF) wall systems, roof and slab insulations, and air leakage requirements.

The results show that applying the energy efficient design scenario to the newly constructed multi-family buildings in the whole US can reduce GHG emissions by 17% and can save up to 69 million metric tons of CO2 (eq) in the whole US. Our analysis of building code evolution showed that a concrete structure built to 2015 standards in Virginia demonstrates a considerable saving and generates 222 metric tons less CO2 (eq) when compared to the same concrete structure built to 2009 standard in North Carolina. Furthermore, the fraction of embodied impacts in a multi-family structure increases from 9.3% to 14.9% after code evolution from IECC 2009 to an energy efficient design scenario.

Moreover, the majority of the energy loads for the buildings including lighting, large appliances, and miscellaneous activities are constant across design scenarios. These loads take up between 56 and 73 percent of the overall energy consumption of a building and this means that energy efficiency improvements mostly influence heating and cooling loads, which suggests that it is more meaningful to report savings caused by design decisions relating to the building envelope in heating and cooling terms, and not in overall energy savings.



10:50am - 11:10am

Buildings as batteries: an experimental investigation into energy efficiency impacts of demand response.

Aditya Keskar1, David Anderson2, Jeremiah Johnson1, Ian Hiskens2, Johanna Mathieu2

1North Carolina State University; 2University of Michigan, Ann Arbor

There is an increasing need for flexible resources to maintain reliable power grid operation due to the combined effect of reduced grid inertia and the addition of supply-side stochasticity caused by renewables. Commercial building heating ventilation and air conditioning (HVAC) systems are attractive candidates for load shifting due to their large thermal inertia and inherently sophisticated building controls. Recent work has suggested potential adverse impacts on energy efficiency associated with such demand response activity.

To explore this phenomenon, we conducted over hundred experiments on three buildings at the University of Michigan and will soon be conducting similar experiments on three buildings at North Carolina State University. We perturb the building temperature setpoints in predefined patterns, causing the building to change its power consumption over and below its baseline power use, thereby acting like a battery from the grid’s perspective. This emulation of energy neutral demand response events is then used to assess the impact of the tests on overall building energy efficiency.

We developed novel metrics to assess the building performance and evaluated the performance of various demand response signal designs. We present results from the experiments and quantify the efficiency of building response by focusing on the round trip efficiency as well as the additional energy consumed by the building while providing this demand response service. The three buildings respond with mean roundtrip efficiencies ranging from 34% to 81%, with individual tests yielding efficiencies far outside that range. We also find that the efficiency of building response depends on the magnitude and polarity of the temperature setpoint changes. Our results are consistent with past experimental results, but inconsistent with past modelling results. At North Carolina State University, we will assess the diversity in building characteristics while evaluating the response of buildings to concurrent demand response events. The change in location of experiments will also help us investigate the impact of local climatic conditions on the efficiency of building response. Our findings offer new and practical insights into the impacts of demand response on building operations and potential challenges needed to be overcome to achieve commercial viability.



11:10am - 11:30am

Do LED lightbulbs save natural gas? Detecting simultaneity bias in examining program impacts

Oluwatobi G. Adekanye, Alex Davis, Inês L. Azevedo

Carnegie Mellon University, United States of America

Reducing energy consumption through energy efficiency has been seen as a cost-effective means of promoting energy reductions in buildings. As a result, different energy efficiency programs have been implemented at the local, state, and federal levels in promoting reductions in building use. Traditional assessments of these programs use ex-ante engineering analyses to estimate these program impacts. However, such analyses will be accurate to the degree that their assumptions are met in the real world, and will be inaccurate when those assumptions are wrong.

Complementing engineering analyses are data-driven approaches, that provide empirical estimates of the energy savings of new technology in specific contexts, such as residential households. With the use of appropriate statistical models, it is possible to empirically determine the impact of these new technologies on energy use. This data-driven approach requires different assumptions about behavior than engineering analyses. One of the most fundamental, and difficult to test, is the assumption that the adoption of new technology occurs in the absence of other changes to a building’s energy profile. Changes that occur at the same time as the introduction of the new technology to the system will lead to biased estimates of the energy savings of the new technology, a phenomenon called simultaneity bias. While the randomized control trial approach may be used to circumvent the possibility of the simultaneity bias (as households are being randomly assigned into treatment and control groups), the high costs and infeasibility for some projects make it difficult to implement. Therefore, most ex-post evaluation studies use other quasi-experimental approaches in estimating program impacts with which the simultaneity bias may be present.

Here, we provide a means of addressing the simultaneity bias by examining monthly electricity and gas billing data from approximately 27,000 households in the City of Palo Alto, California from 2010 to 2016. By using simultaneous measurements of both electricity and gas consumption, we examine the effects of electricity saving programs on gas usage and vice versa. LED lightbulbs, for example, cannot physically change gas usage, therefore finding impacts of LED lightbulb use on gas usage is a first-level indication of simultaneity bias. Using the differences-in-differences and event history model approach, we find varied effects of the different energy efficiency programs with significant average reductions of 3% - 8% for electricity use and 4% - 8% for gas usage. We also find evidence that behavioral programs are more effective than financial incentive programs as energy audit programs show the highest reductions in both electricity and gas use. However, we also find evidence of the simultaneity bias effect, as for some electricity-only programs, we observe significant simultaneous gas reductions. Even after accounting for short and long-run effects, we still find evidence of the simultaneity bias with an LED lightbulb program.

Our findings yield significant implications for future analysis as we find that ex-post evaluation of program impacts needs to be carefully examined such that biases such as the simultaneity bias is eliminated for accurate detection of program impacts.



11:30am - 11:50am

The Potential for Emissions Reductions with Residential Demand Response

Jeremiah Johnson, Madeline Rose Macmillan

North Carolina State University, United States of America

The primary goal of demand response (DR) is to reduce peak electricity demand. In this study, we examine an alternative goal of using DR to reduce air emissions. For the US, we estimate the diurnal and seasonal demand profiles for suitable residential end uses including air conditioning, electric heating, and water heating. We assume that the DR events are load-neutral and test a range of tolerances for demand deferral. We develop an emissions minimization model that utilizes hourly marginal emissions factors for 22 grid regions to show significant potential to reduce CO2 emissions through DR approaches. For each region, we calculate a heating and a cooling coefficient on the ratio of heating or cooling electricity consumption per household to the cumulative annual heating degree day (HDD) and cooling degree day (CDD) values. These coefficients are then applied to region-level hourly HDD or CDD values to estimate hourly heating and cooling electricity demand in each region.

Our results show that the magnitude of the benefits is limited by the length of the demand deferral and DR adoption rate. With participation and high tolerance for load shifting, we estimate up to 15%, 12%, and 13% decreases in CO2 emissions from electric heating, air conditioning, and electric water heating applications, respectively. The potential varies across regions and the regions with high residential demand and high variance in marginal emissions factors yield the greatest potential to reduce CO2 emissions. The magnitude of the emissions reduction can differ greatly from the percent of emissions reduced, as driven by the population of the region and the size of the electric load. Load deferral shows the greatest potential for electric heating during the winter and for air conditioning during the summer months, while electric water heating potential is relatively constant throughout the year. This study shows the magnitude of emissions reduction is sufficiently large and that these emissions reductions can be met without significant loss of energy services. In addition to direct emissions reductions, DR initiatives introduce the possibility of alleviating some of the system constraints that lead to solar and wind curtailment, resulting in greater emissions reductions than DR initiatives alone.