Carnegie Mellon University

MS and Ph.D. Dissertations

Below is a list of MS and PhD theses completed by students in the Green Design Institute.  If interested in copies, please contact the authors directly.

2013 | 2012 | 2011 |2010 | 2009 | 2008 | 2007 | 2006 | 2005 |
2004 | 2001 | 1999 | 19981997 | 19961995

2013 Dissertations

Land Use and Congestion Management Strategies to Promote Urban Environmental Sustainability, Yeganeh Mashayekh, 2013.  Chair: Chris Hendrickson

Reducing greenhouse gas emissions (GHG) is an important social goal to mitigate climate change. A common mitigation paradigm is to consider strategy ‘wedges’ that can be applied to different activities to achieve desired GHG reductions. In this dissertation, I consider a wide range of possible travel demand reduction and traffic congestion management strategies to reduce light-duty vehicle GHG emissions.

To estimate the cost savings associated with the implementation of various travel demand and traffic congestion management strategies, performance measures such as speed, delay, and travel time were assessed for each strategy. These performance measures were then combined with emission factors – amount of pollutants per speed interval – and monetary damage values of each pollutant in terms of mortality, morbidity and environmental damages – dollar per gram of pollutant – to estimate the external environmental cost savings resulting from the implemented strategy. Fuel and time cost savings were simply measured by incorporating the value of time and fuel. Specifically, the external environmental cost of driving in the U.S. including congestion was estimated to be about $110 billion annually. Brownfield developments and LEED certified brownfield developments were assessed as land use and travel demand management strategies to reduce vehicular travel demand. Impacts of these residential developments on vehicle miles traveled (VMT) reduction and the resulting costs (cost of driving time, fuel, and external air pollution costs) were examined. Results show with minimal implementation cost incurred by transportation authorities (about 75-95% less than other VMT reduction measures), both brownfield residential developments and LEED certified brownfield residential developments can be beneficial travel demand strategies, assisting federal, state and local governments with their GHG emissions reduction goals. Compared with conventional developments, residential brownfield developments can reduce VMT and its consequential environmental costs by about 52 and 66 percent respectively. LEED certified residential brownfield developments can have an additional 1% to 12% VMT reduction and a 0.03% to 3.5% GHG reduction compared with conventional developments.

In addition to land use and travel demand management strategies, a number of supply congestion management measures were also assessed. Traffic signal timing and coordination is an effective congestion management strategy. However, not maintaining the timings regularly to assure they respond to vehicle volumes may result in 18 percent increase in the cost of fuel consumed, 13 percent in the cost of travel time and 11 percent in the external environmental costs annually. Other supply management strategies assessed were cases of adaptive traffic control system and high occupancy toll (HOT) lanes. In comparison to one another, while adaptive traffic signal control system results in 7 to 12 percent external environmental cost saving, HOT lanes show zero external environmental cost savings. Driving patterns and speed profiles have significant impacts on the emission of the criteria air pollutants. In some cases, speed improvements resulting from the implementation of a congestion management measure may, in fact, result in the emission of additional criteria air pollutants, thus increasing the external environmental costs. Other interdependencies such as induced demand were also examined. Results show that induced demand from excess capacity resulting from an implementation of a supply congestion management strategy can be significant enough to reduce the benefits gained from the implemented measure in a short period of time.

In addition to analyzing travel demand management, land use changes and congestion management, strategies including fuel and vehicle options and low carbon and renewable power are briefly discussed in this work. I conclude that no one strategy will be sufficient to meet GHG emissions reduction goals to avoid climate change. However, many of these changes have positive combinatorial effects, so the best strategy is to pursue combinations of transportation GHG reduction strategies to meet reduction goals. Agencies need to broaden their agendas to incorporate such combinations in their planning.

Back to Top

How the Timing of Climate Change Policy Affects Infrastructure Turnover in the Electricity Sector: Engineering, Economic and Policy Considerations, Catherine Izard, 2013.  Chairs: Scott Matthews and Chris Hendrickson

The electricity sector is responsible for producing 35% of US greenhouse gas (GHG) emissions. Estimates suggest that ideally, the electricity sector would be responsible for approximately 85% of emissions abatement associated with climate polices such as America’s Clean Energy and Security Act (ACES). This is equivalent to ~50% cumulative emissions reductions below projected cumulative business-as-usual (BAU) emissions. Achieving these levels of emissions reductions will require dramatic changes in the US electricity generating infrastructure: almost all of the fossil-generation fleet will need to be replaced with low-carbon sources and society is likely to have to maintain a high build rate of new capacity for decades. Unfortunately, the inertia in the electricity sector means that there may be physical constraints to the rate at which new electricity generating capacity can be built. Because the build rate of new electricity generating capacity may be limited, the timing of regulation is critical—the longer the U.S. waits to start reducing GHG emissions, the faster the turnover in the electricity sector must occur in order to meet the same target. There is a real, and thus far unexplored, possibility that the U.S. could delay climate change policy implementation for long enough that it becomes infeasible to attain the necessary rate of turnover in the electricity sector.

This dissertation investigates the relationship between climate policy timing and infrastructure turnover in the electricity sector. The goal of the dissertation is to answer the question: How long can we wait before constraints on infrastructure turnover in the electricity sector make achieving our climate goals impossible?

Using the Infrastructure Flow Assessment Model, which was developed in this work, this dissertation shows that delaying climate change policy increases average retirements rates by 200-400%, increases average construction rates by 25-85% and increases maximum construction rates by 50-300%. It also shows that delaying climate policy has little effect on the age of retired plants or the stranded costs associated with premature retirement. In order for the electricity sector to reduce emissions to a level required by ACES while limiting construction rates to within achievable levels, it is necessary to start immediately. Delaying the process of decarbonization means that more abatement will be necessary from other sectors or geoengineering. By not starting emissions abatement early, therefore, the US forfeits its most accessible abatement potential and increases the challenge of climate change mitigation unnecessarily.

Back to Top

Life Cycle Cost and Environmental Implications of U.S. Electric Vehicle and Charging Infrastructure Scenarios, Elizabeth Traut, 2013.  Chair: Jeremy Michalek

This thesis examines life cycle cost, greenhouse gas (GHG) emissions, petroleum use, and policy implications of scenarios for electrified vehicles and charging infrastructure in the U.S., addressing several questions: What mix of vehicles minimizes life cycle cost? GHG emissions? What are the implications of workplace charging in addition to home charging? How much current and potential U.S. residential charging exists? What are the costs and GHG emissions of fast-charging and battery swapping service stations? How sensitive are these results to uncertain parameters? What factors are most critical? and What are the policy implications?

Results indicate that without sufficiently clean electricity, plug-in vehicles (PEVs) with home and workplace charging do not offer substantial reductions in GHG emissions compared to hybrid electric vehicles (HEVs). Benefits improve with low-emission electricity generation. High gas prices ($6/gal) cause PEVs to appear in minimum cost solutions and combined with low vehicle and battery costs (DOE 2030 targets) cause PEVs to dominate.

Currently 79% of households but only 56% of vehicles have home parking where charging could be installed. Excluding renters, who face additional barriers, less than half of U.S. vehicles have reliable access to off-street parking where charging could be installed. This places a major limit on potential penetration of PEVs for the foreseeable future.

Battery swapping stations cost 40% more per vehicle served than fast charging stations without the cost of waiting time during service, but 50% less when it is included. Battery swapping’s cost advantage requires vehicle and battery standardization.

Several policy implications are identified. Gas prices and vehicle and battery prices are identified as price levers to encourage adoption and reduce petroleum consumption, but clean electricity is also needed for GHG emissions reductions. Lack of residential charging could curb adoption and needs attention since parking infrastructure turns over more slowly than the vehicle fleet. With clean electricity, dedicated workplace charging further reduces GHGs. Battery electric vehicle (BEV) adoption is restricted by limited range. Rapid BEV refueling options include fast charging, which incurs costly waiting times during service, or battery swapping, which is faster and potentially less costly but requires vehicle and battery standardization.

Back to Top

2012 Dissertations

Energy Performance Impacts from Competing Low-slope Roofing Choices and Photovoltaic Technologies, Amy Nagengast, 2012.  Chair: Chris Hendrickson

With such a vast quantity of space, commercial low-slope roofs offer significant potential for sustainable roofing technology deployment. Specifically, building energy performance can be improved by installing rooftop energy technologies such as photovoltaic (PV) panels, and/or including designs such as white or green roofs instead of traditional black. This research aims to inform and support roof decisions through quantified energy performance impacts across roof choices and photovoltaic technologies. The primary dataset for this research was measured over a 16 month period (May 24, 2011 to October 13, 2012) from a large field experiment in Pittsburgh, Pennsylvania on top of a commercial warehouse with white, black and green roof sections, each with portions covered by polycrystalline photovoltaic panels. Results from the Pittsburgh experiment were extended to three different cities (San Diego, CA; Huntsville, AL; and Phoenix, AZ) chosen to represent a wide range of irradiance and temperature values.

First, this research evaluated the difference in electricity production from a green-moss roof and black roof underneath photovoltaic panels to determine if the green roof’s cooler air increases the panel efficiency. Second, separate studies examine 1) average hourly heat flux by month for unobstructed and shaded roof membranes 2) heat flux peak time delay, and 3) air temperature across roof types.

Results of this research show green roofs slightly increased (0.8-1.5%) PV panel efficiency in temperatures approximately at or above 250 C (770F) compared to black roofs. However in cool climates, like Pittsburgh, the roof type under the PV panels had little overall impact on PV performance when considering year round temperatures. Instead, roof decisions should place a stronger emphasis on heat flux impacts. The green roof outperformed both black and white roofs at minimizing total conductive heat flux. These heat flow values were used to develop a new, straight-forward methodology to estimate heat flux impacts of different roof types in other climates using ambient temperature and solar irradiance. While managing heat flow is important for building energy performance, roof choices need to include a systems level analysis encompassing a year for the specific region to best quantify the overall energy impacts.

Back to Top

Evaluating Biomass Energy Policy in the Face of Emissions Reductions Uncertainty and Feedstock Supply Risk, Kimberley Mullins, 2012.  Chair: Mike Griffin

Biofuels have received legislative support recently in California’s Low-Carbon Fuel Standard and the Federal Energy Independence and Security Act. Both discuss new fuel types, but neither provides methodological guidelines for dealing with the inherent uncertainty in evaluating their potential life-cycle greenhouse gas emissions. Emissions reductions are based on point estimates only. This work develops a Monte Carlo simulation to estimate life-cycle emissions distributions from ethanol and butanol from corn or switchgrass. Life-cycle emissions distributions for each of the modelled feedstock and fuel pairings span an order of magnitude or more. Corn ethanol emissions range from 50 to 200 g CO2e/MJ, and each feedstock-fuel pathway studied shows some probability of greater emissions than a distribution for gasoline. Potential GHG emissions reductions from displacing fossil fuels with biofuels are difficult to forecast given this high degree of uncertainty in life-cycle emissions. Incorporating uncertainty in the decision making process can illuminate the risks of policy failure (e.g., increased emissions), and a calculated risk of failure due to uncertainty can be used to inform more appropriate reduction targets in future biofuel policies. The current practice of modelling cellulosic biomass yields based on point values that have been aggregated over space and over time conceal important energy supply risks related to depending on biomass for transportation energy, particularly those related to local drought conditions. Using switchgrass as a case study, this work quantifies the variability in expected yields over time and space with a switchgrass growth model and historical weather data. Even with stable, productive states, yields vary from 5 to 20 Mg/ha. Yields are likely to be reduced with increased temperatures and weather variability induced by climate change. Thus, variability needs to be a central part of biomass systems modelling so that risks to energy supplies are acknowledged and risk mitigation strategies or contingency plans are considered. Irrigation, a potential risk mitigation strategy, can very often negate the impacts of drought, although system-wide irrigation is an expensive method to stabilize crops (costing $0.10 to $1.90/gallon). Unless surplus acres of cellulosic crops are planted, there will be insufficient ethanol to meet the EISA targets 10 to 25% of the time under rain-fed conditions. Thinking in terms of yield ranges, not point estimates, is essential in planning a long-term energy system dependent on biomass.

Back to Top

Towards Robust Energy Systems Modeling: Examining Uncertainty in Fossil Fuel-Based Life Cycle Assessment Approaches, Aranya Venkatesh, 2012.  Chairs: Mike Griffin and Paulina Jaramillo

Increasing concerns about the environmental impacts of fossil fuels used in the U.S. transportation and electricity sectors have spurred interest in alternate energy sources, such as natural gas and biofuels. Life cycle assessment (LCA) methods can be used to estimate the environmental impacts of incumbent energy sources and potential im- pact reductions achievable through the use of alternate energy sources. Some recent U.S. climate policies have used the results of LCAs to encourage the use of low carbon fuels to meet future energy demands in the U.S. However, the LCA methods used to estimate potential reductions in environmental impact have some drawbacks. First, the LCAs are predominantly based on deterministic approaches that do not account for any uncertainty inherent in life cycle data and methods. Such methods overstate the accuracy of the point estimate results, which could in turn lead to incorrect and (consequent) expensive decision-making. Second, system boundaries considered by most LCA studies tend to be limited (considered a manifestation of uncertainty in LCA). Although LCAs can estimate the benefits of transitioning to energy systems of lower environmental impact, they may not be able to characterize real world systems perfectly. Improved modeling of energy systems mechanisms can provide more accu- rate representations of reality and define more likely limits on potential environmental impact reductions.

This dissertation quantitatively and qualitatively examines the limitations in LCA studies outlined previously. The first three research chapters address the uncertainty in life cycle greenhouse gas (GHG) emissions associated with petroleum-based fuels, natural gas and coal consumed in the U.S. The uncertainty in life cycle GHG emissions from fossil fuels was found to range between 13 and 18% of their respective mean values. For instance, the 90% confidence interval of the life cycle GHG emissions of average natural gas consumed in the U.S was found to range between −8 to 9% (17%) of the mean value of 66 g CO2e/MJ. Results indicate that uncertainty affects the conclusions of comparative life cycle assessments, especially when differences in average environmental impacts between two competing fuels/products are small. In the final two research chapters of this thesis, system boundary limitations in LCA are addressed. Simplified economic dispatch models are developed to examine changes in regional power plant dispatch that occur when coal power plants are retired and when natural gas prices drop. These models better reflect reality by estimating the order in which existing power plants are dispatched to meet electricity demand based on short-run marginal costs. Results indicate that the reduction in air emissions are lower than suggested by LCA studies, since they generally do not include the complexity of regional electricity grids, predominantly driven by comparative fuel prices. For instance, this study estimates 7–15% reductions in emissions with low natural gas prices. Although this is a significant reduction in itself, it is still lower than the benefits reported in traditional life cycle comparisons of coal and natural gas-based power (close to 50%), mainly due to the effects of plant dispatch.

Back to Top

GHG Emissions and Costs of Developing Biomass Energy in Malaysia: Implications on Energy Security in the Transportation and Electricity Sector, Azman Hassan Mohd Nor, 2012.  Chair: Mike Griffin

Malaysia’s transportation sector accounts for 48% of the country’s total energy use. The country is expected to become a net oil importer by the year 2011. To encourage renewable energy development and relieve the country’s emerging oil dependence, in 2006 the government mandated blending 5% palm-oil biodiesel in petroleum diesel. Malaysia produced 16 million tonnes of palm oil in 2007, mainly for food use. This study addresses maximizing bioenergy use from oil-palm to support Malaysia’s energy initiative while minimizing greenhouse gas emissions from land use change. When converting primary and secondary forests to oil-palm plantations between 270 - 530 g and 120 -190 g CO2 equivalent (CO2-eq) per MJ of biodiesel produced, respectively, is released. However, converting degraded lands results in the capture of between 23 to 85 g CO2-eq per MJ of biodiesel produced. Using various combinations of land types, Malaysia could meet the 5% biodiesel target with a net GHG savings of about 1.03 million tonnes (4.9% of the transportation sector’s diesel emissions) when accounting for the emissions savings from the diesel fuel displaced. Fossil fuels contributed about 93% to Malaysia’s electricity generation mix and emit about 65 million tonnes (Mt) or 36% of the country’s 2010 Greenhouse Gas (GHG) emissions. The government has set a target to install 330 MW biomass electricity by 2015, which is hoped to avoid 1.3 Mt of GHG emissions annually. The availability of seven types of biomass residues in Peninsular Malaysia is estimated based on residues-to-product ratio, recoverability and accessibility factor and other competing uses. It was found that there are approximately 12.2 Mt/yr of residues. Oil-palm residues contribute about 77% to the total availability with rice and forestry residues at 17%. Electricity from biomass can be produced via direct combustion in dedicated power plants or co-fired with coal. The co-firing of the residues at four existing coal plants in Peninsular Malaysia was modeled to minimize cost or GHG emissions. It is found that Malaysia can meet the 330 MW biomass electricity target via co-firing with a cost reduction of about $24 million compared to 100% coal. Optimal GHG reduction for co-firing was found to be 17 Mt lower than 100% coal at a cost of carbon mitigation (COM) of about $22.50/t CO2-eq mitigated. This COM is lower than an implied COM under the newly introduced levy on heavy electricity users in Malaysia.

Gasoline consumed roughly 370 PJ of energy in Malaysia's transportation sector in 2009. Ethanol can be blended with gasoline up to 10% by volume in most vehicles. Peninsular Malaysia's 12.2 Mt/yr of agro-forestry residues can be used for potentially producing 3.8 billion liters ethanol annually. Using a large scale mixed-integer linear optimization, it is found that if Malaysia introduces a 10% ethanol-gasoline blend (E10), approximately 2.9 Mt (24%) of the residues would be used at $5.4 million more cost compared to 100% gasoline (reference case) estimated at $5.2 billion/yr. In the E10 scenario, all cities receive 10% ethanol altogether producing 900 million liters of ethanol. The GHG emissions for 100% gasoline is estimated at 26.4 Mt/yr. The minimum GHG emissions if E10 is implemented in Peninsular Malaysia was found to be 24.5 Mt, 2.0 Mt lower than 100% gasoline, which implies a $4.70/t CO2-eq cost of carbon mitigation (COM). Since only 24% of the available residues are used to produce the E10, the possibility of producing the E10 and electricity via co-firing with coal simultaneously was investigated. This is done by combining the fuel (gasoline/E10) model with the electricity (coal-only/co-firing) model. The costs of the reference case combined scenario (100% gasoline and 100% coal) is estimated at $6.3 billion/yr and emits 63 Mt/yr of GHG emissions. The minimum cost for producing the E10 and co-firing is found to be $30 million lower than the combined reference case. This is achieved by using 5.9 Mt of residues. The minimum GHG emissions level obtained is 17 Mt lower implying a COM of $19.00/t CO2-eq mitigated.

The findings in this research are used to recommend policies for mitigating GHG emissions impacts from the growth of palm oil use in the transportation sector. Policy recommendations are also discussed to ensure a successful implementation of co-firing of biomass and the production of E10 by ensuring a guaranteed supply of residues and financing the high capital cost of the renewable energy program.

Sustainable Transportation and Decision-Making: case studies of transportation decisions to reduce environmental impacts, Rachael Nealer, 2012.  Chairs: Chris Hendrickson and Scott Matthews

Transportation is ubiquitous in the supply chain of producing final goods and services in the United States. Understanding the direct and indirect impacts of freight transportation is particularly important because for most products, the total embodied transportation, the sum of the direct and indirect transportation, is much larger than the direct transportation. It is important for policies to target energy and emissions reductions related to both direct and total embodied transportation.

The primary objectives of this work are to quantify direct and total embodied transportation across sectors by mode, identify potential reductions for total freight energy and emissions, and assess the effectiveness of potential US modal freight policies to reduce energy and emissions in the transportation sector. An additional objective is to demonstrate the feasibility of applying the methodology to a specific sector, specifically food products. This work develops a model based on Input-Output Analysis to estimate total embodied transportation across the supply chain of products and services. The model is applied to analyze potential reductions in mode shift policies and used to compare freight transportation to other life cycle phases of grocery shopping. Overall, rail and truck transportation are the most dominant domestic modes (35% each) for the average sector. Truck transportation requires more energy than rail, and emits about 50% of the domestic freight transportation related emissions. The mode shift and truck efficiency scenarios modeled show energy and emissions reductions of approximately 7% each for the freight transportation sector. When compared to other life cycle phases of grocery shopping, freight transportation emissions are small (10%).

Reductions through mode shift and technology policy may be chosen for various reasons, but applying both could have a larger effect on reducing total freight transportation energy and emissions. Grocery store emissions are highest in food product production and store energy use, not freight or customer transportation. The most effective emissions reductions policies for the grocery shopping system are likely in reducing store energy use. Determining the most effective policies to reduce energy and emissions requires analyzing direct and total energy and emissions, as well as effort for reduction strategies.

Back to Top

2011 Dissertations

Achieving Realistic Energy and Greenhouse Gas Emission Reductions in U.S. Cities, Mike Blackhurst, 2011.  Chair: Scott Matthews

In recognizing that energy markets and greenhouse gas emissions are significantly influences by local factors, this research examines opportunities for achieving realistic energy greenhouse gas emissions from U.S. cities through provisions of more sustainable infrastructure. Greenhouse gas reduction opportunities are examined through the lens of a public program administrator charged with reducing emissions given realistic financial constraints and authority over emissions reductions and energy use. Opportunities are evaluated with respect to traditional public policy metrics, such as benefit-cost analysis, net benefit analysis, and cost-effectiveness.

Section 2 summarizes current practices used to estimate greenhouse gas emissions from communities. I identify improved and alternative emissions inventory techniques such as disaggregating the sectors reported, reporting inventory uncertainty, and aligning inventories with local organizations that could facilitate emissions mitigation. The potential advantages and challenges of supplementing inventories with comparative benchmarks are also discussed. Finally, I highlight the need to integrate growth (population and economic) and business as usual implications (such as changes to electricity supply grids) into climate action planning. I demonstrate how these techniques could improve decision making when planning reductions, help communities set meaningful emission reduction targets, and facilitate CAP implementation and progress monitoring.

Section 3 evaluates the costs and benefits of building energy efficiency are estimated as a means of reducing greenhouse gas emissions in Pittsburgh, PA and Austin, TX. Two policy objectives were evaluated: maximize GHG reductions given initial budget constraints or maximize social savings given target GHG reductions. This approach explicitly evaluates the trade-offs between three primary and often conflicting program design parameters: initial capital constraints, social savings, and GHG reductions. Results suggest uncertainty in local stocks, demands, and efficiency significantly impacts anticipated outcomes. Annual greenhouse gas reductions of 1 ton CO2 eq/capita/yr in Pittsburgh could cost near nothing or over $20 per capita annually. Capital-constrained policies generate slightly less social savings (a present value of a few hundred dollars per capita) than policies that maximize social savings. However, sectors, technologies, and end uses targeted for intervention vary depending on policy objectives and constraints. The optimal efficiency investment strategy for some end uses varies significantly (in excess of 100%) between Pittsburgh and Austin, suggesting that resources and guidance conducted at the national scale may mislead state and local decision-makers. Section 3 then evaluates the impact of rebound effects on modeled efficiency program outcomes. Differential rebound effects across end-uses can change the optimal program design strategy, i.e., the end-uses and technologies targeted for intervention. The rebound effect results suggest that rebound should be integral to effective efficiency program design.

Section 4 evaluates the life cycle assessment costs and benefits of the widespread retrofit of green roofs in a typical urban mixed-use neighborhood. Shadow-cost analysis was used to evaluate the cost-effectiveness of green roofs’ many benefits. Results suggest green roofs are currently not cost effective on a private cost basis, but multi-family and commercial building green roofs are competitive when social benefits are included. Multifamily and commercial green roofs are also competitive alternatives for reducing greenhouse gases and storm water run-off. However, green roofs are not competitive energy conservation techniques. GHG impacts are dominated by the material production and use phases. Energy impacts are dominated by the use phase, with urban heat island (UHI) impacts being an order of magnitude higher than direct building impacts. Results highlight the importance of clarifying sustainable infrastructure costs and benefits across many public and private organizations (e.g., private building owners, storm water agencies, efficiency stakeholders, and roofing contractors) to identify appropriate incentives and effective program design strategies.

Section 5 synthesizes the work and provides guidance for local and state sustainability program administrators. Section 5 highlights the unrealized social benefits associated with sustainability and reflects upon the role of local and state governments in overcoming barriers to achieving more sustainable infrastructure. Section 5 encourages program administrators to consider their local markets for sustainability as influences by resource pricing, weather, infrastructure condition, jurisdiction, and other factors. The differences between sustainability programming and traditional municipal programming are highlighted, namely that sustainability programming often requires self-selection for participation and is subject to new sources of uncertain regarding user behavior, technology breadth and change, and the scope of costs and benefits. These characteristic issues of sustainable infrastructure opportunities provide new challenges to program administrators, requiring new paradigms and support resources.

Section 5 also examines implications for broader policy-making. I apply the building efficiency modeling results to estimate the potential impact of the $11B awarded to states and local governments for clean energy projects as part of the American Recovery and Reinvestment Act of 2009. Assuming all $11B is spent on energy efficiency in residential electricity market, I estimate the $11B would at best achieve a 1% reduction in greenhouse gases over 30 years; however, the reduction is likely much less due to rebound effects and poor decision support. A 30-year 20% reduction in greenhouse gases in the residential electricity market would thus cost at $300 billion, which is about 1% of the total U.S. housing value or nearly the housing value of Dallas, TX.

Section 5 finishes with some recommendations for Federal policy makers and practical guidance for state and local sustainable infrastructure program administrators.

Back to Top

2010 Dissertations

Relating land use and select environmental impacts to U.S. consumption with a focus on agricultural products, Christine Costello, 2010.  Chair: H. Scott Matthews

Approximately 29 percent (%) of the Earth's surface is used to support humanity. Demand for land use is expected to increase by an additional 33% over the next 100 years. Land suitable for crop cultivation is limited, as are pasturelands, both critical for food production. How we use these lands and all land results in environmental impacts. Environmental impacts associated with land use and land use change are many and various. Land use change can cause runoff and sedimentation of soil, contamination of waterbodies with fertilizers and pesticides and release of carbon to the atmosphere. Paving of surfaces causes changes to the hydrology of an area and can create heat island effects. All of these land uses disrupt habitat for other species.

This thesis addresses the following main research questions: What types and how much land occupied by industry and agriculture, in this thesis defined as land in production, are used to meet demands for consumption in the U.S.? What commodities use the most land? How much land is traded between the U.S. and the rest of the world (ROW)? Are there particular categories of land use that dominate supply chains across sectors? How is land use connected to environmental impacts?

First, land in the United States is related to domestic final demand, including land embodied in U.S. exports. To do this an inventory of land with respect to economic sectors is created. Environmentally Extended Input-Output Analysis (EE-IOA) is used to define connections between land in production and consumed goods and services. It is found that agricultural land use is significant in the majority of economic sectors.

Next, the EE-IOA is expanded to include an additional region representative of the rest of world. This enables estimation of the land embodied in U.S. imports. In many land use studies only agricultural products are considered. Through extending this analysis to include the total supply chain land use embodied in the production of good and services it is shown that land use associated with manufactured goods is significant.

Finally, two analyses are executed to demonstrate the connections between land use and environmental impacts. First, Monte Carlo Analysis is employed to approximate the nitrate output within the Mississippi/Atchafalaya River Basin as a result of increased demand for biofuels in the U.S. Nitrate output is related to the formation of hypoxia in the northern Gulf of Mexico. Results indicate that with or without biofuels, our current land use and land use management practices are inadequate. Each year a hypoxic zone forms in the northern Gulf of Mexico the additional land in cultivation will exacerbate this situation. Next, the connection between greenhouse gas (GHG) emissions and land use was explored, again using EE-IOA methods.

Back to Top

Approaches to Reducing the Social Cost of Biofuel Production, Distribution, and Consumption, Matt Kocoloski, 2010.  Chair: Mike Griffin

Biofuels, and specifically next‐generation biofuels such as cellulosic ethanol, have the potential to create economic, environmental, and energy security benefits relative to the fossil fuels that currently power the transportation sector in the United States. However, issues involving ethanol production cost, emissions resulting from land use change, and infrastructure requirements may incur significant social costs. This dissertation examines social costs from different aspects of biofuel production, distribution, and consumption in an effort to inform policies that could reduce these costs.

This dissertation contains seven research chapters that examine social costs of ethanol at different points along the supply chain. This work begins by examining some impacts of cellulosic feedstock production. Land use change, especially indirect land use change, has been the most controversial topic within the biofuel research community in recent years, with some findings indicating that biofuels could be more carbon‐intensive than gasoline. However, cost reductions from cellulosic ethanol could be used to more than offset the increased emissions if policies are in place to balance the impacts. Ethanol production from forest thinnings, on the other hand, could result in a positive externality by reducing wildfire damage while also providing funds for additional fuel treatments.

Decisions regarding cellulosic ethanol facility size and location can have significant impacts on production cost. Cellulosic ethanol refinery investments over the next 12 years are expected to be on the order of $100 billion, so these decisions could be costly if made suboptimally. The rest of the thesis examines costs and impacts of ethanol distribution, promoting a regional fuel strategy that would have ethanol consumed in high‐level blends (such as E85) in regions where it can be produced (mainly the Midwest and Southeast) rather than in low‐level blends throughout the country. Regional distribution would save billions of dollars per year in shipping costs and reduce shipping loads and congestion costs along the rail freight network. Imports of sugarcane ethanol produced in Brazil could be part of this regional fuel strategy, but costs for shipping the fuel from plants to ports within Brazil could be substantial. A key component of this regional fuel strategy is the penetration of both flex‐fuel vehicles and E85 infrastructure throughout ethanol producing regions, but these costs are generally less than the savings from reduced shipping costs.  Next‐generation biofuels such as cellulosic ethanol will play an increasing role in meeting transportation energy demand in the near future. This research will hopefully help shape policies that will allow cellulosic ethanol to meet demand while limiting social cost.

Back to Top

Design Decision Making For Market Systems And Environmental Policy With Vehicle Design Applications, Ching-Shin (Norman) Shiau, 2010.  Chair: Jeremy J. Michalek

The goal of design decision making is to create products to satisfy functional requirements and meet consumer preferences in order to succeed in the marketplace. Using only engineering objectives and constraints may be insufficient to fully describe product performance as market and social objectives are involved in design decisions. This dissertation attempts to address three broad questions at the interface of engineering design, market systems, and public policy in an effort to provide insight for designers, consumers, and policy-makers. 
The first question, “How does market competition affect product design decisions?” is addressed in a game-theoretic framework. Methodology is proposed to account for competitor price reactions to a new product entrant, and the study results indicate that ignoring competitors’ pricing reactions can cause profit overestimation and impede the market performance of a new product design. Furthermore, mathematical analysis shows that consumer preference heterogeneity as a critical factor coupling engineering design and strategic market planning under long-run design competition.

The second question, “What are the economic and environmental implications of plug-in hybrid electric vehicles (PHEVs)?” is addressed through vehicle performance simulation, driver behavior characterization, and life cycle assessment. The results indicate charging between distances matters – PHEVs with small battery packs and short electric travel distances can outperform ordinary hybrid vehicles and large PHEVs if drivers charge frequently. PHEV design and allocation for various social objectives, including minimum petroleum consumption, life cycle costs and greenhouse gas emissions, are analyzed using an optimization framework. The study suggests that alternative PHEV designs are needed for different social targets, carbon allowance policy may have marginal impact to PHEV design under current US grid mix, and the subsidy on battery capacity can be less effective than that on all-electric range because recently developed battery technology allows maximum energy use in the batteries.

The third question addressed is “How does public policy affect vehicle design in a competitive market?” A model integrating vehicle design, oligopolistic market competition, and Corporate Average Fuel Economy (CAFE) regulations is presented to analyze automakers’ design decisions. A distinctive pattern is identified in firms’ vehicle design responses to fuel economy standards at market equilibrium. The results imply that automakers may fail to improve their vehicle fuel economy when a high CAFE standard is imposed. Through a case study of powertrain design incorporating the estimates of recent automotive market data and fuel-efficiency technology options, it is shown that fuel economy design responses are more sensitive to gasoline prices than CAFE policy.

Back to Top

Life Cycle Management of Reducing Impacts on Climate Change at a Regional Level, YuShan (Anny) Huang, 2010.  Chair: Scott Matthews

This dissertation work consists of three studies related to the concept of entities moving beyond the environmental impacts under their direct control and becoming empowered to address the life cycle emissions outside of their physical or political boundaries. Measurements of greenhouse gas (GHG) emissions are guided by carbon footprint protocols, which require estimation of direct emissions (Scope 1) and emissions from direct purchases of energy (Scope 2), but focus less on indirect emissions upstream and downstream of the supply chain (optional Scope 3). The first part of this dissertation uses input-output analysis method to categorize sectors’ indirect upstream supply chains and demonstrated the importance of Scope 3 guideline specificity in improving the performance of footprint accounting. The results show that enterprises can capture a large portion of their total upstream carbon footprint by collecting full emissions information from only a handful of direct suppliers, and Scope 3 footprint capture rates can be improved considerably by sector-specific categorization.

In the second part, the paradoxical “materiality thresholds” for including emission sources in life cycle GHG accounting used by an existing protocol were explored, and the trade-offs between threshold, performance, and efforts in GHG accounting were examined using Structural Path Analysis. It is demonstrated that application of cut-off thresholds results in highly variable performance in footprint capture rate and is not a reliable criterion for including emission sources in GHG footprints.

In the third part, an advanced supply chain decomposition technique was introduced and used in conjunction with a multi-regional input-output model to analyze the life cycle effects of Renewable Portfolio Standard (RPS) and Low Carbon Fuel Standard (LCFS) on reducing the carbon footprint of California enterprises and consumers. This study found that if California achieves the RPS target of 33%, California businesses and consumers can expect to see 2-10% reduction in their carbon footprints. If California achieves a LCFS of 10%, California business and consumers can expect a reduction of 0.5-4% in their total carbon footprints. If a state intends to significantly reduce life cycle carbon footprint of enterprises and consumers through policy interventions, additional policy tools must be considered.

Back to Top

2009 Dissertations

Global Climate Change and Human Behavior: Decreasing Energy Consumption, Shahzeen Attari, 2009.  Chair: Cliff I. Davidson

To decrease carbon dioxide emissions per capita and hopefully reduce the problem of climate change, many scientists have addressed supply-side methods using carbonreduction technologies such as carbon capture and storage. But with increasing population and rising energy demand in developed and developing countries, it is unclear whether supply-side methods alone can make sufficient progress toward solving the problem. This thesis investigates demand-side management methods to facilitate a reduction in carbon emissions. The thesis consists of three main studies. First, I design and implement intervention experiment to facilitate a decrease in energy consumption. Second, I use surveys to understand when and why an individual would accept voluntary actions, soft regulations or hard regulations to curb fossil fuel consumption. Third, I show how lay perceptions of energy consumed by different every-day behaviors differ from actual energy consumption data.

My first experiment was an eight-week intervention study that examined whether holding people accountable for their behaviors leads to energy conservation (n=100). The intervention asked participants for reasons why they did or did not engage in energy conserving behaviors, with questions focusing on household operations, transportation, and food purchases. Results show that the intervention, in general, did not facilitate behavior change in these sectors. However, an important finding is that participants erroneously perceived there is not much difference in energy saved by several different behaviors. Furthermore, 60% of participants perceived a change in their own behavior over the course of the study even though no overall behavior change occurred. This result could imply that participants have optimistic illusions regarding their own behavior change.

My second experiment was a study on preferences to change behavior. Pittsburgh residents (n = 209) reported their preferences for voluntary actions, soft regulations, and hard regulations to (a) limit the number of SUVs and trucks and (b) increase green energy use for household energy consumption. These two goals were presented in one of two motivating frames, as addressing either environmental or national security issues. For thegoal of limiting SUVs and trucks, results indicated that participants favored voluntary actions over hard regulations, and soft regulations over voluntary actions. For the goal of increasing green energy, results indicated that participants preferred both voluntary actions and soft regulations over hard regulations, but had no significant preference between voluntary actions and soft regulations. Participants’ environmental attitudes (as assessed using the New Ecological Paradigm scale) had a strong positive relationship with support for regulatory strategies intended to change the behaviors in question. Women were more likely to support voluntary actions than men. The loss of personal freedom was frequently mentioned as a reason for saying no to hard regulations.

My third experiment studied how participants (n=505) perceive energy consumption and savings for household, transportation, and recycling behaviors. Participants’ showed a tendency to overestimate energy consumption and savings for low-energy behaviors and underestimate energy consumption and savings for high-energy behaviors. On average, participants underestimated the amount of energy used or saved by different behaviors. Pro-environmental attitudes and higher numeracy scores were associated with more accurate perceptions of energy consumption. However, participants who reported engaging in a greater number of environmental behaviors had less accurate perceptions of energy consumption. On average, participants reported that engaging in energy conserving behaviors would not be difficult for any of the behaviors considered.

Back to Top

Coal Supply and Cost Under Technological and Environmental Uncertainty, Melissa Chan, 2009.  Chair: Granger Morgan

This thesis estimates available coal resources, recoverability, mining costs, environmental impacts, and environmental control costs for the United States under technological and environmental uncertainty. It argues for a comprehensive, well-planned research program that will resolve resource uncertainty, and innovate new technologies to improve recovery and environmental performance. A stochastic process and cost (constant 2005$) model for longwall, continuous, and surface mines based on current technology and mining practice data was constructed. It estimates production and cost ranges within 5 – 11 percent of 2006 prices and production rates. The model was applied to the National Coal Resource Assessment. Assuming the cheapest mining method is chosen to extract coal, 250 – 320 billion tons are recoverable. Two-thirds to all coal resource can be mined at a cost less than $4/mmBTU. If U.S. coal demand substantially increases, as projected by alternate Energy Information Administration (EIA), resources might not last more than 100 years. By scheduling cost to meet EIA projected demand, estimated cost uncertainty increases over time. It costs less than $15/ton to mine in the first 10 years of a 100 year time period, $10-$30/ton in the following 50 years, and $15-$90/ton thereafter.

Environmental impacts assessed are subsidence from underground mines, surface mine pit area, erosion, acid mine drainage, air pollutant and methane emissions. The analysis reveals that environmental impacts are significant and increasing as coal demand increases. Control technologies recommended to reduce these impacts are backfilling underground mines, surface pit reclamation, substitution of robotic underground mining systems for surface pit mining, soil replacement for erosion, placing barriers between exposed coal and the elements to avoid acid formation, and coalbed methane development to avoid methane emissions during mining. The costs to apply these technologies to meet more stringent environmental regulation scenarios are estimated. The results show that the cost of meeting these regulatory scenarios could increase mining costs two to six times the business as usual cost, which could significantly affect the cost of coal-powered electricity generation.

This thesis provides a first estimate of resource availability, mining cost, and environmental impact assessment and cost analysis. Available resource is not completely reported, so the available estimate is lower than actual resource. Mining costs are optimized, so provide a low estimate of potential costs. Environmental impact estimates are on the high end of potential impact that may be incurred because it is assumed that impact is unavoidable. Control costs vary. Estimated cost to control subsidence and surface mine pit impacts are suitable estimates of the cost to reduce land impacts. Erosion control and robotic mining system costs are lower, and methane and acid mine drainage control costs are higher, than they may be in the case that these impacts must be reduced.

Back to Top  

Life Cycle Greenhouse Gas Emissions, Consumptive Water Use And Levelized Costs Of Unconventional Oil In North America, Aweewan Mangmeechai, 2009.

Conventional petroleum production in many countries that supply U.S. crude oil as well as domestic production has declined in recent years. Expanding the U.S. energy mix to include oil sands and oil shale may be an important component in diversifying and securing the U.S. energy supply. At the same time, life cycle GHG emissions of these energy sources and consumptive water use are a concern. The goal of this study is to determine the life cycle greenhouse gas (GHG) emissions and consumptive water use of synthetic crude oil (SCO) derived from Canadian oil sands and U.S. oil shale to be compared with U.S. domestic crude oil. Levelized costs of SCO derived from Canadian oil sands and U.S. oil shale were also estimated.

The results of this study suggest that CTL with no carbon capture and sequestration (CCS) and current electricity grid mix is the worst while crude oil imported from United Kingdom is the best in GHG emissions. The life cycle GHG emissions of oil shale surface mining, oil shale in-situ process, oil sands surface mining, and oil sands in-situ process are 43% to 62%, 13% to 32%, 5% to 22%, and 11% to 13% higher than those of U.S. domestic crude oil.  Oil shale in-situ process has the largest consumptive water use among alternative fuels, not including biofuel, evaluated due to consumptive water use in electricity generation. Life cycle consumptive water use of oil sands in-situ process is the lowest. Specifically, fresh water consumption in the production processes is the most concern given its scarcity. Given current information, it is inconclusive whether unconventional oil would require more or less consumptive fresh water use than U.S. domestic crude oil production. It depends on the water conservative strategy applied in each process.

Increasing import of SCO derived from Canadian oil sands and U.S. oil shale would slightly increase life cycle GHG emissions of the U.S. petroleum status quo. The expected additional 2 million bpd of Canadian SCO from oil sands and U.S. oil shale would increase life cycle GHG emissions of the U.S. petroleum status quo on average only 10 and 40 kg CO2 equiv/bbl, or about 7.5 and 29 million tons CO2 equiv/year. However this increase represents less than 1 and 5% of U.S. transportation emissions in 2007.

Because U.S. oil shale resources are located in areas experiencing water scarcity, methods to manage the issue were explored. The result also shows that trading water rights between Upper and Lower Colorado River basin and transporting synthetic crude shale oil to refinery elsewhere is the best scenario for life cycle GHG emissions and consumptive water use of U.S. oil shale production. GHG emissions and costs of water supply system contribute only 1-2% of life cycle GHG emissions and 1-6% of total levelized costs. The levelized costs of using SCO from oil shale as feedstock are greater than SCO from oil sands, and CTL. The levelized costs of producing liquid fuel (gasoline and diesel) using SCO derived from Canadian oil sands as feedstock are approximately $0.80-$1.00/gal of liquid fuel.The levelized costs of SCO derived from oil shale are $1.6 -$4.5/gal of liquid fuel (oil shale surface mining process) and $1.6-$5.2/gal of liquid fuel (oil shale in-situ process).  From a GHG emissions and consumptive water use perspective, CTL requires less consumptive water use than oil shale in-situ process but produces more GHG emissions than oil shale in-situ and surface mining process, unless CTL plant performs CCS and renewable electricity.

Back to Top

A Life-Cycle Approach to Technology, Infrastructure and Climate Policy Decision-Making: Transitioning to Plug-In Hybrid Electric Vehicles and Low-Carbon Electricity, Constantine Samaras, 2009

In order to mitigate the most severe effects of climate change, large global reductions in the current levels of anthropogenic greenhouse gas (GHG) emissions are required in this century to stabilize atmospheric carbon dioxide (CO2) concentrations at less than double pre-industrial levels. The Intergovernmental Panel on Climate Change (IPCC) fourth assessment report states that GHG emissions should be reduced to 50-80% of 2000 levels by 2050 to increase the likelihood of stabilizing atmospheric carbon dioxide (CO2) concentrations. In order to achieve the large GHG reductions by 2050 recommended by the IPCC, a fundamental shift and evolution will be required in the energy system.
Because the electric power and transportation sectors represent the largest GHG emissions sources in the United States, a unique opportunity for coupling these systems via electrified transportation could achieve synergistic environmental (GHG emissions reductions) and energy security (petroleum displacement) benefits. Plug-in hybrid electric vehicles (PHEVs), which use electricity from the grid to power a portion of travel, could play a major role in reducing greenhouse gas emissions from the transport sector. However, because GHG emissions from PHEVs depend on the electricity source that is used to charge the battery, meaningful GHG emissions reductions with PHEVs are conditional on low-carbon electricity sources. Power plants and their associated GHGs are long-lived, so decisions made regarding new electricity supplies within the next ten years will affect the potential of PHEVs to play a role in a low-carbon future in the coming decades. This thesis investigates the life cycle engineering, economic, and policy decisions involved in transitioning to PHEVs and low-carbon electricity.
The government has a vast array of policy options to promote low-carbon technologies, some of which have proven to be more successful than others. This analysis uses a hybrid life cycle assessment to evaluate options and opportunities for large GHG reductions from plug-in hybrids. After the options and uncertainties are framed, this work uses engineering economic analysis to evaluate the actions required for adoption of PHEVs at scale and the implications for low-carbon electricity investments. This work concludes with an examination of what lessons can be learned for climate, innovation, and low-carbon energy policies from the evolution of wind power from an emerging alternative energy technology to a utility-scale power source. Policies to promote PHEVs can take lessons learned from the successes and challenges of wind powers development to optimize low-carbon energy policy and going forward.
Back to Top  

2008 Dissertations

Alternative Transportation Fuels: Infrastructure Requirements and Environmental Impacts for Ethanol and Hydrogen, Heather Wakeley, 2008.  Chairs: Chris Hendrickson and Scott Matthews

This dissertation evaluates infrastructure requirements for ethanol and hydrogen as alternative fuels. It begins with an economic case study for ethanol and hydrogen in Iowa. A large-scale linear optimization model is developed to estimate average transportation distances and costs for nationwide ethanol production and distribution systems. Environmental impacts of transportation in the ethanol life cycle are calculated using the Economic Input-Output Life Cycle Assessment (EIO-LCA) model. An EIO-LCA Hybrid method is developed to evaluate impacts of future fuel production technologies. This method is used to estimate emissions for hydrogen production and distribution pathways.
Results from the ethanol analyses indicate that the ethanol transportation cost component is significant and is the most variable. Costs for ethanol sold in the Midwest, near primary production centers, are estimated to be comparable to or lower than gasoline costs. Along with a wide range of transportation costs, environmental impacts for ethanol range over three orders of magnitude, depending on the transport required. As a result, intensive ethanol use should be encouraged near ethanol production areas.
Fossil fuels are likely to remain the primary feedstock sources for hydrogen production in the near- and mid-term. Costs and environmental impacts of hydrogen produced from natural gas and transported by pipeline are comparable to gasoline. However, capital costs are prohibitive and a significant increase in natural gas demand will likely raise both prices and import quantities. There is an added challenge of developing hydrogen fuel cell vehicles at costs comparable to conventional vehicles.
Two models developed in this thesis have proven useful for evaluating alternative fuels. The linear programming models provide representative estimates of distribution distances for regional fuel use, and thus can be used to estimate costs and environmental impacts. The EIO-LCA Hybrid method is useful for estimating emissions from hydrogen production. This model includes upstream impacts in the LCA, and has the benefit of a lower time and data requirements than a process-based LCA.  
 Back to Top 

Trade, Consumption, and Climate Change: An Input-Output Study of the United States, Christopher Weber, 2008.  Chair: Scott Matthews

The previous decade of rapid globalization has brought substantial changes to the U.S. economy, and at the same time, increasing concern about global climate change has brought it to the forefront of U.S. environmental policy. Climate change, like most environmental problems, can be traced back through complex pathways to consumer demand for goods and services, and globalization has further increased the physical and mental distance between consumption and the environmental consequences of production. This work brings together concepts of international trade, consumption, and climate change to analyze the climate impacts of American consumption and trade. 
Two broad analyses, one at the economy level and one at the individual household level, are followed by two case studies on specific commodities: food and electronics.We find that a rising proportion (13-30% in 2004) of the CO2 impacts of U.S. consumption are taking place outside of the geographical boundaries of the U.S., and importantly, that the proportion taking place outside the U.S. is increasingly taking place in countries which have not adopted strong climate policies, with around 75% of emissions in US imports taking place in non-Annex B parties to the Kyoto Protocol. This is especially true for electronics, where consumption is consistently increasing, product turnover is very high, and production is particularly concentrated in a small set of developing nations. Food, despite consumer worries about the safety and wisdom of its increasingly global supply chains, remains mostly sourced in the US, though a high-impact category for climate change nonetheless, due both to high transportation and production impacts.
International trade has several policy consequences for climate change—it has been proposed as both a problem, via “leakage” of emissions to developing countries, and a potential solution through coercive or cooperative avenues. Coercive policies such as carbon tariffs, though seemingly an effective way to protect competitiveness and increase participation in global climate frameworks, have several disadvantages including missing a large share of emissions embodied in trade and potentially being both illegal and ineffective due to counteractions. Utilizing global supply chains and trade through cooperative actions such as tariff reduction, clean development programs, and technologytransfer could solve many of the political and policy issues of coercive actions and climate policy in general.
Back to Top  

2007 Dissertations 

A Mixed-Unit Input-Output Model For Life Cycle Assessment: Development, Uncertainty, And Application, Troy Hawkins, 2007.  Chair: Scott Matthews

Material flows of heavy metals and other toxic chemicals are of great concern for companies, regulators, researchers and society at large. Understanding how and why we use these toxic chemicals can help us to use them more efficiently. In this thesis I develop a mixed-unit, input-output life-cycle assessment model (MUIO-LCA) capable of tracking cadmium, lead, nickel, and zinc flows among roughly 500 sectors of the U.S. economy.
Flows of cadmium, lead, nickel and zinc commodities are modeled using data provided by the U.S. Geological Survey (USGS) by adding sectors to national input-output accounts developed by the U.S. Bureau of Economic Analysis (B.E.A.). First an aggregate, summary level model based on the B.E.A. 12 by 12 sector monetary input-output tables is introduced and used to explain the method. Then a detailed model is constructed by augmenting the detailed B.E.A. 1997 Benchmark Input-Output make and use tables. The 1997 Benchmark Input-Output Accounts consist of 483 commodities, 491 industries, and 13 final demand sectors. Flows of cadmium, lead, nickel, and zinc were added as an additional 46 commodities, 20 industries, and 10 final demand sectors.
Flows of an additional 103 materials are modeled with use factors developed from U.S.G.S. production and consumption data. These materials include 48 metals, 38 minerals, coal combustion by-products, agricultural products, and wood products. An assessment of the uncertainty in model is presented. This includes a discussion of the error in input-output life-cycle assessment models as well as an analysis of uncertainty in model parameters. Because of the large data requirements for the creation of the mixed-unit make and use tables, definition of distributions for each data point was not possible. Uncertainty in the use factors for the additional 103 materials was estimated by calculating these factors using consumption data as well as production data for the period 1997 to 2004.
To demonstrate MUIO-LCA the material use associated with the production of an average automobile is modeled. Results are compared to those of a comprehensive, process-based inventory of the material content and supply chain material use for a generic U.S. family sedan performed by the USAMP LCA Project ('98). Results for the lead content of an average car were compared to an inventory provided by the Clean Car Campaign (Trumble '98). This case study highlighted the need for care in interpreting the results of the MUIO-LCA model. MUIO-LCA provides information about commodities consumed directly and throughout the supply chain of a product. While these values provide guidance on the possible material content of a product itself, they also include the tangential use of materials consumed in processes but not included in products.  An exploratory examination of the material intensity of the 483 monetary transaction commodities in the MUIO-LCA model is performed using the total requirements matrix. Results are summarized for the top 10 sectors for material use of metal commodities included in the MUIO-LCA model as well as the 103 additional materials.
Troy Hawkins
trh (at)
Back to Top

A Life Cycle Comparison of Coal and Natural Gas for Electricity Generation and the Production of Transportation Fuels, Paulina Jaramillo, 2007. Chair: Scott Matthews and Mike Griffin

Demand for electricity is expected to increase in the next 25 years. Currently, 50% of the electricity generated in the U.S. is produced using coal. Although natural gas has traditionally been used by the commercial, industrial and residential sector, demand for natural gas for electricity generation has increased in the past decade and this growth is expected to continue in the next 25 years. Since demand is growing but North American supply is expected to remain constant, alternative sources of natural gas will need to be developed. LNG has been identified as one alternative, and plans to increase imports of this fuel are underway. In addition, synthetic natural gas could be produced from coal to meet some of the increasing demand for natural gas.
The demand for natural gas by the transportation sector is currently negligible, but worldwide interest on natural gas-derived transportation fuels (such as natural gas based Fischer-Tropsh Liquids and Compressed Natural Gas) is increasing. The U.S. could either produce these fuels internally, requiring larger imports of LNG, or import them from natural gas-rich countries. Alternatively, the U.S. could produce transportation fuels from coal. Although non-existent in 2005, by 2030 coal-to-liquid-fuel producers are expected to consume as much coal as coke plants. Thus, the production of transportation fuels is an additional end-use where coal and natural gas could compete as the fuel of choice.
The goal of this research is to compare coal and natural gas for use by the electric power sector and for the production of transportation fuels in the next 25 years. This comparison concentrates on the life cycle GHG emissions of these fuels. In addition to comparing natural gas and coal to determine which fuel is better suited for each end-use, a comparison of each end-use will also be performed in order to help determine which is a better use of each fuel.  Two main results arise from this research. First, it was found that in a future whereadvanced power plant technologies with carbon capture and sequestration are used, coal and globally sourced natural gas could have very similar life cycle GHG emissions. This begs the question of whether investing billions of dollars in LNG/SNG infrastructure will lock us into an undesirable energy path that could make future energy decisions costlier than ever expected and increase the environmental burden from our energy infrastructure. Second, it was found that the use of transportation fuels derived from coal and natural gas will not help the U.S. reduce the GHG emissions associated with the life cycle of transportation fuels, and in a worse case scenario, the use of these alternative fuels could in fact increase these GHG emissions. In addition, it was found that there is high uncertainty associated with the energy security benefits that could be associated with the consumption of transportation fuels derived from coal.
Paulina Jaramillo
Post-doctoral researcher
Department of Civil and Environmental Engineering
Carnegie Mellon University
pjaramil (at)
Back to Top

Quantifying and Assessing the Impacts of Heavy Metal Flows: Fate, Transport, and Impacts of Lead Use in U.S. Product Manufacturing, Cortney Higgins, 2007.  Chair: Scott Matthews

Heavy metals, such as lead, are toxic, yet despite awareness of this toxic- ity, they are used throughout the global economy. Eliminating their use would reduce risk, but they have useful physical and chemical properties and suit- able substitutes have proven difficult to find. For example, due to previous investigations of lead's toxicity, it was removed as a gasoline additive in the US and other countries beginning in the 1980s. However every automobile and truck shipped contains a lead-acid battery, and many other products still contain lead. For decision makers, being able to better link the flows and uses of lead with the potential human health and ecotoxicity impacts of its release is an important goal.
This dissertation seeks to contribute to this need by developing methods to dynamically track lead flows throughout the economy, and also to link the flows with eventual impact. This is accomplished by the creation of various mixed-unit input-output models of lead and lead compounds in the US economy, at various levels of aggregation (from 12 to approximately 100 sectors). For the latter, a dynamic lead flow model was created to aid in tracking the flows of lead to various media over a 15 year timeframe, 1990 to 2004. The United States' Toxics Release Inventory (TRI) tracks releases to various media of both lead and lead compounds, and this data is a key input into the methods used. These models allow decision makers to recognize which sectors are the sources of the greatest direct and total lead and lead compound emissions.
Beyond tracking lead flows, this dissertation also links lead flows with impacts. To this end, impact assessment values for lead and lead compounds from two well known impact assessment methods, CalTox and the U.S. Environmental Protection Agency's Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts (TRACI) model, were used.  The results of the impact assessment suggest that the parameters of CalTOX, as used in TRACI, may lead to underestimates of the actual concentra- tions of lead in the environment. Thus decision makers focused on lead and other heavy metals should explore other impact assessment methods to ensure that the impacts of these substances are better understood.
Back to Top

Greening Construction Processes with an Input-Output-Based Hybrid Life Cycle Assessment Model, Aurora Sharrard, 2007.  Chair: Scott Matthews

This research uses a life cycle assessment (LCA) framework to create a more specific and accurate estimate of the environmental impacts of construction processes. The construction industry is spatially and economically diverse, yet fragmented because it is always changing, occurring everywhere, and dependent upon a variety of specialized contractors. Because the construction industry is so complex, modeling traditional construction processes is required in order to understand the specific environmental implications of these activities; this goal would be best achieved with the process LCA approach, though process LCAs are data intensive and time consuming given the “made to order” nature of each construction project. Conversely, the input-output LCA approach allows for a more inclusive view of the construction industry (i.e., not just traditional on-site construction activities), and can be used to identify which processes should be of principal concern. A combination of both approaches allows the strengths of each to be used to the best advantage. Consequently, a hybrid LCA framework produces a comprehensive analysis that includes the economy-wide effects of construction while addressing specific on-site construction activities.
The input-output-based hybrid LCA framework selected for this research is based on Carnegie Mellon University’s (CMU) Economic Input-Output Life Cycle Assessment (EIO-LCA) modeling tool. The input-output-based hybrid was created by combining a new “Hybrid” feature that uses the EIO-LCA interface and updated and reformulated environmental effects vectors for EIO-LCA’s thirteen construction sectors. The final stage of this research models a variety of construction case studies on the input-output-based hybrid LCA framework to demonstrate its broad applicability.
The hybrid LCA model for construction processes detailed here is designed help decision-makers make more informed decisions regarding the construction industry, adding environmental quality and sustainable development as goals instead of unintentional benefits. The model’s focus on construction site and ancillary support activities helps identify opportunities for improvement in the construction industry that may otherwise be missed and provides a holistic assessment that identifies priority areas for future research into the environmental impacts of construction processes.
Aurora L. Sharrard, PhD
Research Manager
Green Building Alliance
333 East Carson St., Suite 331
Pittsburgh, PA 15219
aurorasharrard (at)
Back to Top

An Electricity-focused Economic Input-output Model: Life-cycle Assessment and Policy Implications of Future Electricity Generation Scenarios,Joe Marriott, 2007.  Chair: Scott Matthews

The electricity industry is extremely important to both our economy and our environment: we would like to examine the economic, environmental and policy implications of both future electricity technologies and the interaction of this industry with the rest of the economy. However, the tools which currently exist to analyze the potential impacts are either too complex or too aggregated to provide this type of information.
Because of its importance, and the surprising lack of associated detail in the inputoutput model of the U.S. economy, the power generation sector is an excellent candidate for disaggregation. This work builds upon an existing economic inputoutput tool, by adding detail about the electricity industry, specifically by differentiating among the various functions of the sector, and the different means of generating power. We build a flexible framework for creating new industry sectors, supply chains and emission factors for the generation, transmission and distribution portions of the electricity industry. In addition, a systematic method for creating updated state level and sector generation mixes is developed.
The results of the analysis show that the generation assets in a region have a large impact on the environmental impacts associated with electricity consumption, and that interstate trading tends to make the differences smaller. The results also show that most sector mixes are very close to the U.S. average due to geographic dispersion of industries, but that some sectors are different, and they tend to be important raw material extraction or primary manufacturing industries. Further, in scenarios of the present and future, for electricity and for particular products, results show environmental impacts split up by generation type, and with full supply chain detail. For analyses of the current electricity system and products, economic and environmental results match well with external verification sources, but for analyses of the future, there is significant uncertainty. Future work in this area must address the inherent uncertainty of using an economic model to generate emissions values, although the framework of the model allows for infinite expansion and adjustment of assumptions.
Joe Marriott (at)
Back to Top

2006 Dissertations

U.S. Biomass Energy: An Assessment of Costs & Infrastructure for Alternative Uses of Biomass Energy Crops as an Energy Feedstock,William Morrow, 2006.  Chair:  Mike Griffin and Scott Matthews

Reduction of the negative environmental and human health externalities resulting from both the electricity and transportation sectors can be achieved through technologies such as clean coal, natural gas, nuclear, hydro, wind, and solar photovoltaic technologies for electricity; reformulated gasoline and other fossil fuels, hydrogen, and electrical options for transportation. Negative externalities can also be reduced through demand reductions and efficiency improvements in both sectors. However, most of these options come with cost increases for two primary reasons: (1) most environmental and human health consequences have historically and are currently excluded from energy prices; (2) fossil energy markets have been optimizing costs for over 100 years and thus have achieved dramatic cost savings over time. Comparing the benefits and costs of alternatives requires understanding of the tradeoffs associated with competing technology and lifestyle choices.
Bioenergy advocates propose its use as an alternative energy resource for electricity generation and transportation fuel production, primarily focusing on ethanol. These advocates argue that bioenergy offers environmental and economic benefits over current fossil energy use in each of these two sectors as well as in the U.S. agriculture sector. However, estimates of bioenergy resource reveal that bioenergy is only capable of offsetting a portion of current fossil consumption in each sector. As bioenergy is proposed as a large-scale feedstock within the United States, a question of “best use” of bioenergy becomes important. Unfortunately, bioenergy research has offered very few comparisons of these two alternative uses. This thesis helps fill this gap.
This thesis compares the economics of bioenergy utilization by a method for estimating total financial costs for each proposed bioenergy use. Locations for potential feedstocks and bio-processing facilities (co-firing switchgrass and coal in existing coal fired power plants and new ethanol refineries) are estimated and linear programs are developed to estimate large-scale transportation infrastructure costs for each sector. Each linear program minimizes required bioenergy distribution and infrastructure costs. Truck and rail are the only two transportation modes allowed as they are the most likely bioenergy transportation modes. Switchgrass is chosen as a single bioenergy feedstock. All resulting costs are presented in units which reflect current energy markets price norms (¢/kWh, $/gal). The use of a common metric, carbon-dioxide emissions, allows a comparison of the two proposed uses. Additional analysis is provided to address aspects of each proposed use which are not reflected by a carbon-dioxide reduction metric. Using switchgrass as an electricity generation feedstock offers more than twice the amount of carbon-dioxide emission reductions as using switchgrass as an ethanol feedstock (370 versus 160 million short tons per year respectively; representing 14% and 12% of electricity and transportation sector annual CO2 emissions). Total costs, including capital, labor, feedstock, and transportation, is more certain for electricity production than for ethanol; 20 - 45 $/ton CO2 mitigated versus free - 80 $/ton CO2 mitigated respectively. In both cases, mitigation cost is a variable of fossil energy costs. Coal price are very stable as compared to crude oil prices and therefore, more risk is inherent in ethanol economics than in electricity economics.
Additional analysis comparing life-cycle benefits and burdens though full-cost accounting methods also favors bioenergy for electricity production. Agricultural impacts are neutral, while criteria pollutants increase with ethanol use and decrease with bioenergy electricity production. Moreover, ethanol use could cause an increase in groundwater toxicity, a risk that is not associated with electricity production. Considering other available alternative technologies, switchgrass co-firing in existing coal power plants is the least costs retrofitting option available to existing coal fired power plants wishing to lower their carbon emissions. Plug hybrids offer increased system efficiencies over current gasoline-propulsion systems, thereby lowering criteria pollutants and greenhouse gas emissions all at a cost less than or comparable to ethanol. However, shifting transportation energy demands into the United States’ antiquated electrical grid will require large-scale electricity infrastructure investments. The economic impact of a large-scale transfer of energy from petroleum to electricity should be a topic of future research.

Back to Top

2005 Dissertations

Technology Change and Environmental Management for Cement Manufacturing:The Cement Industry in the United States (2004-2050) Jose Luis Aguirre, 2005.  Chair: Chris Hendrickson

Historically, the cement industry has been challengedwith the requirement of improving its manufacturing process while reducing itsfootprint on the environment.  Atthe same time, global competition poses more challenges to improving the bottomline of the business. Research and development of pollution abatementtechnology for cement manufacturing is key for effectively operating in this newenvironment. These new technological advancements compete against establishedtechnologies when cement manufacturers evaluate different pollution preventionstrategies.
This research developed a quantitative tool tobenchmark various technologies available to produce Portland cement in theUnited States. The model “Technology Change Evaluation for the Cement Industry”(TCECI) was developed to achieve this goal considering a full cost approach.Several production scenarios were designed and evaluated to represent the currentand potential future conditions of the cement industry in the United States.The decision making process to select the Best Available Technology (BAT) forcement manufacturing in the United States considered the minimization of theprivate and the total cost (i.e., including private and social costs) underdifferent multi-pollutant approaches. One of these approaches considered theminimization of carbon dioxide emissions from the calcination of raw materialsand the combustion of the fuel from cement manufacturing. These emissions wereestimated for each production scenario considering an emission tax scheme andan emission allowance trading program.
The most relevant result obtained from thisresearch is the integration of environmental and social aspects of cementmaking into the current decision making process for technology change. Thisintegration led to production alternatives with improved environmental, socialand economic performance. Additionally, the results of this research indicatethat the current technology mix for cement manufacturing in the United Stateslimits the feasibility of new cement plants when considering the full costapproach. However, the results of the analysis indicate that the implementationof BAT in existing plants (under the conditions and characteristics assumed bythe TCECI model) improves their overall economic and environmental performance.The reduction estimated for the full cost ranged from 19% to 22% whilecomparing the baseline scenario for the year 2004 with a  multipollutant approach (i.e., in 2004dollars per ton of clinker, $50 –production scenario No. 8- and $48 –productionscenario No. 7- versus $61 from the baseline scenario).
Finally, the results also indicate that within thelimited sample of production scenarios considered there is large variability inthe estimated uncertainty of the costs associated with the production ofcement, the air emissions reported from the production process and theperformance data from available technologies for pollution control and processoptimization. The differences of the social costs estimated for each productionscenario were found statistically more significant when considering the effectof the use of alternative fuels (i.e., tire fuel instead of coal) than the effectof a more stringent regulatory environment.
Since performance data for control technologies andair emissions are becoming more important to private and public policy decisionmaking, it is recommended that the Environmental Protection Agency and the cementindustry treat uncertainty explicitly, by means of adopting standardizedmeasurement and reporting methodologies for air emissions among other relevantmeasures.
Jose Luis Aguirre
jlag04 (at)
Back to Top

Future Electricity Generation: An Economic and Environmental Life Cycle Perspective on Near-, Mid- and Long-Term Technology Options and Policy Implications, Joule Bergerson 2005.  Chair: Lester Lave

The U.S. electricity industry is currently experiencing and adapting to enormous change including concerns related to security, reliability, increasing demand, aging infrastructure, competition and environmental impacts. Decisions that are made over the next decade will be critical in determining how economically and environmentally sustainable the industry will be in the next 50 to 100 years. For this reason, it is imperative to look at investment and policy decisions from a holistic perspective, i.e., considering various time horizons, the technical constraints within the system and the environmental impacts of each technology and policy option from an economic and environmental life cycle perspective.
This thesis evaluates the cost and environmental tradeoffs of current and future electricity generation options from a life cycle perspective. Policy and technology options are considered for each critical time horizon (near-, mid-, and long-term). The framework developed for this analysis is a hybrid life cycle analysis which integrates several models and frameworks including process and input-output life cycle analysis, an integrated environmental control model, social costing, forecasting and future energy scenario analysis.

The near-term analysis shows that several recent LCA studies of electricity options have contributed to our understanding of the technologies available and their relative environmental impacts. Several promising options could satisfy our electricity demands. Other options remain unproven or too costly to encourage investment in the near term but show promise for future use (e.g. photovoltaic, fuel cells). Public concerns could impede the use of some desirable technologies (e.g. hydro, nuclear). Finally, less tangible issues such as intermittency of some renewable technologies, social equity and visual and land use impacts, while difficult to quantify, must be considered in the investment decision process.
Coal is a particularly important fuel to consider in the U.S. and is the main focus of this thesis. A hybrid life cycle analysis including the use of process level data, Economic Input-Output Life Cycle Assessment (EIO-LCA) and the Integrated Environmental Control Model (IECM) quantify a range of potential impacts for new power plants. This method provides a more complete and consistent basis for comparing different technologies. While Integrated Coal Gasification Combined Cycle (IGCC) technology has clear environmental benefits for bituminous coals over conventional pulverized coal plants, the advantages are less clear for the lower ranked coals at present. Near-term implementation of this technology is hampered by concerns about its reliability and performance. A full scale U.S. installation of this technology would settle the performance concerns while more stringent emissions standards would increase its value. In the mid-term analysis, this thesis explores alternative methods for transport of coal energy. A hybrid life cycle analysis is critical for evaluating the cost, efficiency and environmental tradeoffs of the entire system. If a small amount of additional coal is to be shipped, current rail infrastructure should be used where possible. If entirely new infrastructure is required, the mine mouth generation options are cheaper but have increased environmental impact due to the increased generation required to compensate for transmission line losses. Gasifying the coal to produce methane also shows promise in terms of lowering environmental emissions.
The long-term analysis focuses on the implications of a high coal use future. This scenario analysis focuses on life cycle issues and considers various generation and control technologies. When advanced technologies such as gasification with carbon capture and sequestration are used, emissions during generation decrease to a level where environmental discharges from extraction, processing and transportation become the dominant concern. The location of coal, coal composition and mining method are important in determining the overall impacts.  Coal is an inherently dirty fuel. However, for the next half century, coal is likely to play a major role in electricity generation. In deciding how much coal to use, the U.S. must understand the cost and environmental implications of the technologies available, including the whole life cycle of the fuel and the facilities used from extraction, transport, generation, and use or disposal of by products.

Dr. Joule A. Bergerson
University of Calgary
Chemical and Petroleum Engineering
2500 University Drive NW, Room 602
Earth Sciences Building
Calgary, Alberta T2N 1N4
jbergers (at)
Back to Top 

Regional economic input-output analysis-based life-cycle assessment, Gyorgyi Cicas, 2005.  Chair: Chris Hendrickson

Economic input-output analysis-based life-cycle assessment (e.g., the EIO-LCA model by Carnegie Mellon University) has been used by academia, industry and government to estimate the energy and fuel use, toxic and non-toxic emissions and wastes associated with products and services in the United States ( Since its introduction in the mid-1990s, I-O based LCA has been adopted by other countries as well, and the development of national, monetary and physical I-O tables-based models is under way. However, the EIO-LCA model uses national average economic and environmental data and many decisions would be wise to take into account local or regional characteristics. Thus regionalized models would be useful for decision-making on regional and local levels. This dissertation introduces a regional economic input-output analysis-based LCA (REIO-LCA) model which is based on the Gross State Product (GSP) and state level environmental data. The national input-output matrix is modified by using regional economic multipliers and the regional matrices are linked to energy (electricity and fuels), emission factors, and hazardous waste generation factors. The use of the model is demonstrated through case studies.

Back to Top

Life Cycle Assessment of Residential Buildings, Luis Ochoa, 2005.  Chair: Chris Hendrickson

Residential building construction represented about 4.2% ofthe US Gross Domestic Product in 2000, and residences consumed nearly 20% oftotal US energy consumption. However, design and construction of residential buildings is often notconducted with an analysis of the life cycle costs and environmental impacts.  In this paper, we outline an approachto a comprehensive life cycle analysis for residences, using the results of atypical construction cost estimate to map into tools for environmental lifecycle assessment (using the Carnegie Mellon economic input-output life cycleassessment model) and for resources required during the use phase of residences(using the DOE Energy Saver model). In essence, material costs are mapped into input-output sectors and theEIO-LCA model applied to assess environmental impacts.  Similarly, operating inputs such aselectricity or natural gas are estimated from the Home Energy Saver model andmapped into EIO-LCA sectors.  Theresult of using our toolset is a full life cycle assessment based upon theconstruction cost estimate. 
Luis Ochoa
luis1a1 (at)
Back to Top

2004 Dissertations

Using Life Cycle Assessment to Inform Nanotechnology Research and Development, Shannon Lloyd, 2004.  Chair: Lester Lave

By reducing the energy and materials required to provide goods and services, nanotechnology has the potential to provide more appealing products while improving environmental performance and sustainability. However, while nanotechnology offers great potential, it is unlikely to be the first entirely benign technology. My hypothesis is that a technological push towards greater investment in nanotechnology without a commensurate consideration of the net environmental benefits will inevitably lead to cases where the nanotechnology substitute is inferior to the product or process replaced. Whether and how soon the promise of improved environmental quality could be realized depends on phrasing life cycle questions during research and development and pursuing commercialization intelligently.
Efforts have been initiated to develop a fundamental understanding of the behavior of nanotechnology-based materials in natural systems and their influence on biological systems. This understanding will improve the ability to project the direct environmental and health effects of these materials. To obtain a complete picture, it is also necessary to consider life cycle and sustainability implications on nanotechnology-based products. I present a framework employing quantitative analysis to evaluate projected nanotechnology-based products. I use technology scenarios and prospective hybrid life cycle assessment to estimate the economic and environmental life cycle implications of two projected nanotechnology-based products. In the case of using nanocomposites in light-duty vehicle body panels, the ability to disperse nanoscale particles in polymers would reduce vehicle weight thereby improving fuel economy. In the case of nanofabricated catalysts, the ability to position and stabilize platinum-group metal particles in automotive catalyst would reduce the amount of platinum-group metal required to meet emissions standards thereby reducing mining and refining activities. For each application, I compare a conventional product to its nanotechnology-based substitute to assess whether the nanotechnology substitute can be cost-effective and improve environmental quality.
Life cycle assessment is typically used to estimate the resource and environmental implications associated with existing products. Changing a product to reduce its environmental impact after the product has been developed can cost orders of magnitude more than making the change during research and development. As shown here, policy makers and industry can identify technology scenarios and employ prospective life cycle assessment during early research and development to evaluate future products and emerging technologies. The ability to evaluate life cycle implications of alternative courses of action during research and development improves the ability to evaluate tradeoffs, optimize products for all aspects of life cycle performance, and make more strategic R&D choices. A more informed understanding of the commercial, societal, and technological possibilities and its consequences will enable better decisions in regards to the selection, development, and commercialization of nanotechnology.
Back to Top

Two Essays on Problems of Deregulated Electricity Markets, Dmitri Perekhodtsev, 2004, Chair: Lester Lave

1. The data from California energy crisis of 2000 suggests that the largest departures of observed electricity prices from the estimates of the competitive price occur when demand approaches market capacity. This paper studies models of unilateral and collusive market power applicable to electricity markets. Both suggest a unique mechanism explaining the increase of the price-cost margin with demand. The empirical test of these models provides more evidence for unilateral market power than for behavior suggesting tacit collusion.

2. In order to preserve the stability of electricity supply, electric generators have to provide ancillary services in addition to energy production. Hydro generators are believed to be the most efficient source of ancillary services because of their good dynamic flexibility. This paper studies optimal operation decisions for river dams and pumped storage facilities operating in markets for energy and ancillary services as well as the change in the water shadow price in presence of ancillary services markets. The analysis is applied to valuation of the ancillary services provided by hydro resources in the Tennessee Valley Authority. A simulation of ancillary services markets shows that TVA’s hydro resources providing ancillary services can allow for substantial savings in total costs of energy provision. Optimal hydro scheduling in markets for energy and ancillary services increases the value of TVA’s hydro resources by 9% on average and up to 26% for particular units. As a result of hydro participation in ancillary services markets water shadow prices of river dams drop significantly allowing for tightening hydro constraints in favor of other water uses.

Dr. Dmitri Perekhodtsev
6 Canal Park
Suite 708
Cambridge, MA 02140
DPerekhodtsev (at)
Back to Top

2001 Dissertations

Use of enterprise resource planning systems for life cycle assessment and product stewardship, Antje Januschkowetz, 2001.  Chair: Chris Hendrickson

This research shows how enterprise resource planning (ERP) systems can be utilized for environmental life cycle assessments (LCA) of products and product stewardship (PS). Substantial effort has been focused on standardizing, simplifying and developing LCA. However, LCA studies continue to be expensive. Regulations in Europe require PS data, such as detailed bills of materials and substances in weights. Many companies are committed to improve their environmental performance, but they are hindered by the lack of information. ERP systems include data that can be utilized for LCA as well as for PS.

Companies use ERP systems to plan, organize, direct, measure, and control their logistics, financial and human resources. So far, no environmental management functionality has been included in ERP systems. Current ERP systems include materials only partially registered in weights, data registration being process-, but not product-oriented, and data registered on highly aggregated levels, such as the building or plant level. Energy and water use, as well as emissions are not integrated. Environmental reporting functionality is not included.

In the thesis, a model for the use of ERP systems to perform life cycle inventories and obtain PS data is developed. The implementation is demonstrated on an existing SAP R/3 system, the world most widely used ERP system. The following information was integrated: weights in bill of materials, links between equipment and materials, waste data based on an equipment basis or calculated from differences of gross and net weight, energy and water consumption data related to equipment, and allocation procedures from higher reporting levels to products. The thesis describes how a life cycle inventory on a product can be compiled using the enhanced SAP R/3 system. Allocation methods for supplies on a product level are demonstrated. As an example, the thesis demonstrates how an LCI of an automotive component was performed. In addition, the thesis discusses synergies with environmental management systems. In a cost effectiveness analysis, it is found that current costs of stand-alone solutions used for data registration and reporting for environmental management systems and LCA can be decreased significantly using the developed SAP R/3 system.

Back to Top

Assessment and design of industrial environmental management systems, Deanna Matthews, 2001.  Chair: Chris Hendrickson

Environmental management systems (EMS) are gaining acceptance as a tool for organizations to monitor activities with environmental impacts. Current EMS frameworks, such as the International Organization for Standardization EMS (ISO 14000), call for organizations to establish an environmental policy, set goals and targets for current environmental impacts, develop procedures and practices for tasks that have environmental impact, and conduct regular audits and reviews of the system. However, these EMS frameworks focus on the process of identifying and monitoring environmental impacts. The EMS do not provide information to management decision makers that can be used to evaluate options to cost-effectively reduce environmental impacts.

This research evaluates current EMS frameworks from an organizational perspective. A comparison of firms in the automotive assembly sector shows that facilities certified to ISO 14000 have a higher toxic waste per vehicle ratio and are just as likely to be out of compliance with air emission permits as facilities which have not been certified. Limitations of the ISO 14000 EMS are identified by applying the standard to several industrial scenarios which have environmental impacts. The exercise demonstrates the lack of comprehensive data collection and dissemination in the existing EMS structure. The ISO 14000 EMS also does not share effective characteristics of other environmental regulations and initiatives. An examination of safety management, however, reveals that non-financial factors can be improved when part of organizational strategy. Common outcome measures of safety are widely used by organizations, government agencies, and insurers to identify poor performance and target operational changes.

Environmental management systems must be redefined to be more effective for organizations. Effectiveness should be assessed with respect to regulatory compliance, changes in environmental impact, and cost. A model EMS is proposed that is based on providing information linked to each of these aspects. An essential component of EMS should be relevant, useful, and timely outcome measures. Outcome measures can focus the EMS as a tool to improve management decisions to reduce environmental impacts and to reduce costs. This EMS model can be implemented across a wide range of industries using the existing data collection and analysis systems of organizations.

Back to Top

Environmental performance benchmarking of manufacturing plants in Mexico and the United States, Ruth Reyna-Caamano, 2001.  Chair: Chris Hendrickson and Lester Lave

This research compares the environmental performance of manufacturing plants in Mexico and the United States. To do this, I develop an environmental performance benchmarking method in terms of a quantitative instrument, the environmental performance indicator (EPI), which measures the ratio of various air emissions to the production for each manufacturing plant. Benchmarking environmental performance method for manufacturing plants involves six steps. The first step is gathering the environmental and economic data of manufacturing plants in Mexico and the U.S. from company and government sources. The second step involves the analysis and standardization of the environmental and economic data from each country. The third step involves the selection of the industrial sectors to be studied. In this study, the rubber tire industry, Portland cement industry, and coal-fired electricity generation industry were the sectors selected. The fourth step is to develop the EPI's. Selection of the output measure is a main concern in this step. The fifth step is to benchmark the manufacturing plants through the quantitative EPI's. The sixth step involves looking at aspects that affect the analysis of the environmental performance at a plant level. Results from this thesis research show that environmental performance of manufacturing plants in Mexico and the U.S. can be compared using the EPI. The results shows that nitrogen oxide (NO x ) emissions of Mexican rubber tire plants are lower than U.S. rubber tire plants but sulfur dioxide (SO2 ), and particulate matter (PM) emissions are almost three times higher. In addition, Mexican Portland cement plants have better environmental performance than U.S. Portland cement plants because Mexican plants have the newest Portland cement technology in comparison with the U.S. plants. Along the U.S.-Mexico border, the environmental performance of the two Mexican coal-fired power plants that were studied, show up worse than the U.S. coal-fired power plants, because U.S. power plants are burning a cleaner fuel. In addition, the technology of the two Mexican plants did not feature emission reduction provisions, that could have been included as part of the original design.

Back to Top

1999 Dissertations

Development of user centered environmental software systems, Octavio Juarez, 1999.  Chair: Chris Hendrickson, Jim Garrett

This dissertation describes a framework that has been developed to specify, design, implement, and evaluate user-centered environmental software tools. The addition of visualization techniques to perform data analysis was explored in this framework. Life-cycle assessment (LCA) was used as a case study to illustrate the framework, which has several elements: (1) A method was developed to study a problem domain, which yielded the user tasks, the data models, and the visualization needs for that domain. (2) The user interface and the visualizations added to the environmental software are designed using a method that starts by using the task models and the data models produced by the domain analysis. (3) Finally, a method to evaluate the environmental software performance was developed. The system evaluation includes the time used to perform the task and the solution quality. The effect of adding more software components to the software system was explored based on the parameters of time and quality. It was found that the information visualization component improved the data analysis performance. Quality of results and time to perform the task were good indicators of how information visualization helped to improve the environmental system performance. However, the component added to help users in the problem formulation had no effect on the performance. Many suggestions to improve the software component during problem formulation were derived from the analysis of data produced in the experiment.

Back to Top

The external costs of air pollution and the environmental impact of the consumer in the United States economy, H. Scott Matthews, 1999.  Chair: Lester Lave

Despite many years of environmental regulation, significant levels of air pollution are generated by the provision of goods and services in the United States. In 1992, approximately $180 billion of such damages occurred in the U.S.

To reduce environmental damage, analysis tools such as Life Cycle Assessment (LCA) have been developed to better understand the total impacts of products and processes. However, these methods are hindered by the boundary problems and circularities that exist. We propose a solution to the boundary problem, a Leontief input-output (IO) model augmented by environmental impact information to determine the direct and total supply chain effects resulting from the production of the 500 commodity sectors contained in the Department of Commerce's 1992 IO table. The model considers environmental impacts externally to the basic Leontief framework. We have generated a substantial data set linking releases of criteria pollutants and greenhouse gases with manufacturing activities in each sector.

The total air pollution releases found for each commodity are combined with a range of environmental damage valuation studies to estimate the external costs of these activities. Our results include the consideration of indirect (supply chain) effects as opposed to simply the direct effects from producing commodities. The production of electricity generates 34 cents of external costs per dollar cost, but the average commodity generates less than 4 cents.

The results are also combined with the Consumer and Producer Price Indices as well as the Consumer Expenditure Survey to determine the external costs associated with buying and selling commodities. The average dollar spent by consumers generates about 3 cents of external cost, while producers generate 5 to 9 cents. The results show that the average American household's spending generates roughly 40 tons of carbon dioxide equivalent releases per year.

Finally, using our data set of current emissions, we consider the effects of various policies to reduce emissions, including command and control and market-based initiatives. Market-based initiatives are projected to save billions of dollars in expenditures if enacted for sulfur dioxide, nitrogen oxides, and volatile organic compounds. In addition, if new regulations were set to reduce external costs, significant improvements over current levels would result.

Back to Top

1998 Dissertations

Comprehensive product life-cycle analysis using input output techniques, Satish Joshi, 1998.  Chair: Lester Lave

Effective environmentally conscious decision making requires information about environmental consequences of alternative materials, designs, manufacturing processes, product use patterns and disposal. Life cycle assessment (LCA) is a systematic tool to provide this information. LCA attempts to trace out the major stages and processes involved over the entire life cycle of a product covering: raw materials acquisition, manufacturing, product use, recycling and disposal, quantifying the environmental burdens at each stage. The goal of LCA is to facilitate a systems view in product and process evaluation.

The LCA approach is widely recognized as a useful framework and attempts are underway to integrate life cycle thinking into business decisions. A major international initiative in this direction is the series of environmental management standards (EMS) proposed by the International Standards Organization, widely known as ISO 14000. Standards being developed for inclusion under ISO 14000 cover principles and guidelines for conducting LCA for product evaluation. Similarly, the EPA document "Guidance on acquisition of environmentally preferable products and services" prepared to help implement the President's Executive order 12873, recommends LCA approach in all federal procurement. Many eco-labeling and product take back regulations in Europe require life-cycle environmental analysis of products.

The current methods for LCA, proposed by the Society for Environmental Toxicology and Chemistry and the USEPA, suffer from problems of arbitrary boundary definition, inflexibility, high cost, and data confidentiality.

This dissertation develops alternative analytical models, databases and software to conduct quick, cheap and yet comprehensive LCAs. The core of the analytical model consists of the 518 sector economic input-output tables for the US economy augmented with various sector level environmental impact vectors. The environmental impacts covered include, global warming, acidification, energy use, non-renewable ores consumption, eutrophication, conventional pollutant emissions and toxic releases to the environment. Several alternative models are proposed for environmental assessment of individual products and processes using the aggregate input-output data. To demonstrate the method, a case study comparing the life cycle environmental performance of steel and plastic automobile fuel tank systems is presented.

Back to Top

Design and analysis of product takeback systems: an application to power tools, Markus Klausner, 1998.  Chair: Chris Hendrickson

Product takeback has lately been receiving increased attention among industry and policy makers. Unsustainable waste-management practices for post-consumer products have inspired regulatory agencies to impose responsibility for end-of-life products on their manufacturers as they are the entities most capable of influencing end-of-life costs by product design. As product takeback does not belong to a manufacturing firm's core competencies and opportunities to reclaim value from end-of-life products have not yet been explored, most of today's takeback programs are operated at a loss if takeback costs cannot be passed on directly to consumers. This applies to a voluntary power tool takeback program in Germany. Focusing on this example, this thesis provides a systems approach to the analysis and design of product takeback systems. The economic and environmental implications of power tool takeback are studied using detailed manufacturer information, data empirically gathered by the author, and information collected from literature. Based on the conclusion that the current takeback system is environmentally beneficial but not economically sustainable, an alternative product takeback system for power tools is devised. The alternative takeback system would result in a net profit from product takeback, competitive advantages, and a large return volume. Key characteristics of the developed power tool takeback system include the combination of profitable remanufacturing with unprofitable materials recycling, the inclusion of repair returns in remanufacturing, buy-back of end-of-life power tools, and an automated classification of returned products based on electronic data logging during the use stage. The latter is based on sensor-based measurements of customer use patterns for product takeback and is widely applicable to electromechanical and electronic products. This technology was conceptualized by the author and his colleagues at Bosch and then implemented by researchers at Bosch. In this thesis, it is described and applied to the reuse of electric motors of consumer products as a move towards high levels of product recovery. An analysis proves that motor reuse is not only technically feasible but also associated with a large cost savings potential even though the initial manufacturing costs increase due to the implementation of sensor-based data recording.

Back to Top

Materials management and recycling for nickel-cadmium batteries, Rebecca Lankey, 1998.  

Rechargeable battery use is expected to continue growing with the increasing prevalence of portable electronics, appliances, and tools. Nickel-cadmium batteries (NiCds) are widely used in these applications and represent a large volume of toxic materials in common use. While the management of lead-acid batteries is well-established, that of NiCds is in the formative stage.

This thesis considers the relationship between material flows of lead, cadmium, and nickel used in batteries. Battery materials, characteristics, and recycling technologies for NiCd, nickel metal-hydride, and lithium-based batteries are considered. The definition of an environmentally friendly battery is developed, as are substitution models. Life cycle emissions for NiCds are characterized, focusing on materials management during recycling. A materials mass balance of a facility recycling NiCds is constructed. Using publicly-available information, it is difficult to account for all outputs of the metals studied.

The life cycle energy use of NiCds is estimated. Manufacturing batteries using some recycled materials instead of all virgin materials uses about half the energy. Economic issues are presented, and potential costs and revenues for recycling facilities are estimated. The potential net revenue from NiCd recycling depends on current metal prices. The consumer NiCd recycling rate is estimated, and a standardized equation for calculating the total NiCd recycling rate is proposed. NiCd processing facilities are modeled using distillation software, and an example tutorial is developed.

Battery regulations and initiatives are summarized for the U.S. and other countries. U.S. legislation has encouraged the development of a nationwide collection system for NiCds. In Europe bans on the use of cadmium are being considered, although other battery types also use toxic and hazardous materials.

When the total life cycle emissions, energy, and waste generation for each battery type are considered, allowing niche battery markets to develop and promoting recycling for all battery types may be more environmentally friendly than simply banning specific battery types. This thesis provides a basis for continuing evaluation of NiCds and for the analysis of emerging rechargeable battery technologies which have not been comprehensively studied.

Back to Top

Life cycle models of conventional and alternative-fueled automobiles, Heather MacLean, 1998.  Chair: Francis McMichael and Lester Lave

This thesis reports life cycle inventories of internal combustion engine automobiles with feasible near term fuel/engine combinations. These combinations include unleaded gasoline, California Phase 2 Reformulated Gasoline, alcohol and gasoline blends (85 percent methanol or ethanol combined with 15 percent gasoline), and compressed natural gas in spark ignition direct and indirect injection engines. Additionally, I consider neat methanol and neat ethanol in spark ignition direct injection engines and diesel fuel in compression ignition direct and indirect injection engines.

I investigate the potential of the above options to have a lower environmental impact than conventional gasoline-fueled automobiles, while still retaining comparable pricing and consumer benefits. More broadly, the objective is to assess whether the use of any of the alternative systems will help to lead to the goal of a more sustainable personal transportation system. The principal tool is the Economic Input-Output Life Cycle Analysis model which includes inventories of economic data, environmental discharges, and resource use. I develop a life cycle assessment framework to assemble the array of data generated by the model into three aggregate assessment parameters; economics, externalities, and vehicle attributes.

The first step is to develop a set of 'comparable cars' with the alternative fuel/engine combinations, based on characteristics of a conventional 1998 gasoline-fueled Ford Taurus sedan, the baseline vehicle for the analyses. I calculate the assessment parameters assuming that these comparable cars can attain the potential thermal efficiencies estimated by experts for each fuel/engine combination. To a first approximation, there are no significant differences in the assessment parameters for the vehicle manufacture, service, fixed costs, and the end-of-life for any of the options. However, there are differences in the vehicle operation life cycle components and the state of technology development for the combinations.

Overall, none of the alternatives emerges as a clear winner, lowering the externalities and improving sustainability, while considering technology issues and vehicle attributes. The majority of the alternatives are not likely to displace the baseline automobile. However, the attractiveness of the alternatives depends on the focus of future regulations, government priorities, and technology development.

If long-term global sustainability is the principal concern, then improvements in fuel economy alone will not provide the level of reduction in impact required. A switch to renewable fuels (e.g., alcohols or diesel produced from biomass) to power the vehicles will likely be necessary. (Abstract shortened by UMI.)

Back to Top

1997 Dissertations

Estimation of environmental implications of construction materials and designs using life cycle assessment techniques, Arpad Horvath, 1997.  Chair: Chris Hendrickson

The concepts of sustainable development and "environmental friendliness" have lately become very important to government and industry. Increasing public and regulatory pressures have inspired varying degrees of activity. The construction industry has paid particular attention to the ecological impacts of new facilities, energy use of buildings, asbestos contamination and removal, indoor air quality, and several other important environmental issues. Yet many impacts relative to constructed facilities' life cycle (planning, design, construction, occupancy/operation, and demolition) have not been systematically and comprehensively studied. Engineers and architects, as well as owners, planners, operators and other construction professionals should understand the environmental and energy implications of construction materials, processes and alternative facility designs because their everyday decisions carry substantial social implications. The objective of this thesis is to provide information, metrics and analysis tools related to the environmental implications of construction materials and alternative design choices in construction. In particular, this thesis compares in three case studies the environmental effects of asphalt and reinforced cement concrete pavements, steel and reinforced concrete bridge girders, and wood and light-gauge steel framing for residential applications. Systems-level analysis is performed using Life Cycle Assessment/Analysis (LCA) techniques. A primary tool used in the analysis is the Economic Input-Output LCA (EIO-LCA) method developed by Carnegie Mellon University's Green Design Initiative research group. Additional data and information, primarily for the use and disposal life cycle stages, have been collected from literature. The inputs of material production quantified include the use of fuel, electricity, ores and fertilizers, and the environmental emissions (outputs) quantified include toxic chemical releases, ozone depletion potential, conventional air pollutant emissions, and hazardous waste generation and management.

The results indicate that determining which alternative in a comparison is better solely on environmental terms can be difficult. Uncertainties in the data used in the analysis call for careful interpretation of the results. Differences in the longevity of design alternatives may make one alternative "environmentally friendlier" when the impacts are annualized. The manufacturing life cycle stage has larger environmental effects for the three case studies than the use and disposal stages. For asphalt and cement concrete pavements, asphalt appears to be a more environmentally sustainable selection due to its environmentally beneficial high recycling rates. Reinforced concrete girders for bridges appear to have lower overall environmental burdens than steel girders. The study of wood framing vs. light-gauge steel framing for homes indicates that wood framing has lower environmental impacts.

Back to Top

1996 Dissertations

Regional integrated municpal solid waste management in the Northeastern United States, Garrick Louis, 1996.  Chair: Francis McMichael

This thesis evaluates the extent to which regional integrated municipal solid waste management (RIMSWM) eliminates the gap in MSWM capacity the Northeastern U.S. It also investigates how regulation, administration, markets and technology (RAMT) have impacted MSWM nationwide. Three definitions of a region are used to assess the impacts of RIMSWM. Regional mass balances are employed to calculate the ratio of available to needed MSWM capacity in a region. This ratio is defined as the percent capacity. It is used to measure a region's capacity gap. A dataset of annual gate receipts in tons per year was compiled for all the operating MSW landfills and incinerators in the 12-State region. These gate receipts constituted available operating capacity for the host region. The regional mass balance results show that RIMSWM does not eliminate the capacity gap for Counties or States. However, it does so for the 12-States when they recycle. Thus the Federal policy of RIMSWM can work across multiple States provided that interstate MSW transfers are not restricted.

Historical factors are shown to have fragmented MSWM into a locally implemented set of unit operations. Regulations are shown to have contributed to private regional facilities. The impact of the Carbone decision on local flow control is discussed. Proposed Federal legislation to empower State restrictions on the interstate transportation of MSW is analyzed. The waste management industry is shown to have three major revenue earners which accounted for 52% of the industry's revenue in 1994. New opportunities are revealed for shipping companies under the policy of reverse logistics. Recycling is shown to be the most costly waste management option. The thesis closes with a recommendation to update and expand the dataset created for the analysis. A flexible inventory system for municipal recyclables is also proposed.

Back to Top

Measuring environmental performance: sources and implications of uncertainty, Tse-Sung Wu, 1996.  Chair: Francis McMichael

This thesis is a study on the sources and implications of measurement uncertainty in the estimation of toxic industrial emissions. Currently, there is no explicit treatment of uncertainty in measuring and calculating industrial plants' toxic discharges. Yet, in implementing the normative environmental goal of minimizing the dispersion of toxic materials (e.g., pollution prevention, design for the environment, and emissions trading, using analytical tools such as life cycle assessment), existing toxic emission inventories are being used for purposes much beyond those for which they were originally intended. To the extent that more public policy and management decisions are based on toxic emission data, the quality of these data becomes critical.

This work analyzes the existing EPA's Toxic Release Inventory (TRI) data regarding the methods by which surveyed plants estimated their toxic releases. The TRI is a collection of environmental discharges and transfers reported by some 25,000 industrial facilities in the US of more than 300 toxic chemical substances listed by the EPA. In addition, case studies of industrial plants were conducted to ascertain their discharge estimate methodologies, to estimate the uncertainties of their toxic discharge data, and to describe what kinds of management decisions, if any, are supported by these data.

The research reveals that within the small opportunity sample of case studies, there is great variability in the estimated uncertainty of emissions estimates; how plants treat the reporting of TRI data; how plants calculate discharges; in the decisions that are supported. Some interviewed plants face a non-linear loss function regarding decisions that are supported by toxic emissions data. Furthermore, it finds that the current reporting of TRI data is insufficient for explicitly reporting data uncertainty, which can lead to sub-optimal firm and public policy decisions. Finally, in an analysis using toxicity data, this research finds that from 1988 to 1993, while sources of toxic emissions decreased their overall emissions, the toxicity of the discharges from a significant number increased.

Policy recommendations include treating uncertainty more explicitly to make datasets like the TRI more useful for more precise design for the environment applications.

Back to Top

1995 Dissertations

Integrated material selection for the environment (ImSelect): A systematic methodology and computer prototype development for engineering material selection with consideration of life cycle environmental burdens, Rosy Chen, 1995. 

Traditionally, product design has been driven by marketing, economic, engineering, and manufacturing constraints. The key for environmentally conscious design is to integrate environmental considerations into traditional design process and making environmentally oriented decisions in product design.

In this dissertation, a systematic methodology which incorporates the environmental life cycle burdens into current material selection process is developed. This methodology provides a systematic approach that links environmental considerations with engineering analysis at the stage of material selection. This methodology integrates product performance requirements, shape constraints, material properties, the manufacturing process, environmental burdens and costs and provides a framework for comprehensive consideration of engineering and environment aspects related to material selection. A computer prototype is developed by implementing the methodology presented in this research. This computer prototype can be used as a decision support tool for material selection with environmental considerations. The generic method and computer prototype then are demonstrated by a case study of midsize computer housing, by comparing candidate materials--2 types of plastics, aluminum, and electro-galvanized steel. While there are many uncertainties in the data for the example considered, electro-galvanized steel emerges the choice for this type of PC housing with the lowest total product cost including environmental cost.

This research indicates that environmental cost does affect the material selection. Depending on the material energy or toxic release may predominate the life cycle environmental cost. Weighting environmental attributes, e.g., energy consumption and toxic release the same for all materials hides the true environmental impact of different materials. Priority for pollution prevention policies--public or private--should target different environmental attributes depending on the material.

This work also shows that it is not a good method to compare life cycle costs of material per unit weight because different materials require different weight for the same design criteria. Integrating LCA with engineering analysis can yield better results than the approach based on environmental burdens per unit weight. This is because that the weight of material and fabrication processes required for the specific design criteria are determined based on the engineering calculations or on the designer's experience and incorporated into environmental life cycle analysis simultaneously.

Back to Top