Organizational learning and knowledge transfer are key elements within any firm when considering the firm’s competitive advantage and long-term goals. Yet, the roles of learning and knowledge transfer in a multi-product production setting are not well understood. Production and operations management literature suggests production of a variety of products is largely harmful, yet the organizational learning literature suggests there may be benefits to heterogeneity. This work explores the significance of a multi-product environment on organizational learning and knowledge transfer by studying a US-owned overseas manufacturing facility that is a leading producer of high technology hardware components. The firm produces 5 generations of high-volume focus products as well as a collection of non-focus products [an assortment of small volume products related to the focus products]. We draw on 10 years of firm archival data and qualitative data collected to shed insights into how different levels of product mix (5 generations of a focus product, thousands of minor variations on products to meet customer specifications, and an assortment of small volume products related to the focus product) impact organizational learning differently and why knowledge transfers across some products and not others by examining the role that processes play in these product transitions. Our results reconcile differences between the organizational learning and production and operations management literatures by finding support for both advantages and disadvantages to product mix on the production line depending on the extent of product differences. We find that short-term productivity improves with bounded diversity – specifically, when multiple generations of the same product are produced in the same facility. This positive impact on productivity of having multiple generations of the same product on the line may in part be explained by the firm’s ability to successfully transfer knowledge from older to newer generations of the product, improving long-term productivity, though we find benefits for focus product heterogeneity over and above the benefits from knowledge transfer. In contrast, we find short-term productivity is decreased when the production line is faced with variety across products that are too different from each other (e.g. different form factors) and across minor product variations (i.e. customer-specific product variations).

Electricity generated from wind power is both variable and uncertain. Wind forecasts provide valuable information for wind farm management, but they are not perfect. Chapter 2 presents a model of a wind farm with compressed air energy storage (CAES) participating freely in the day-ahead electricity market without the benefit of a renewable portfolio standard or production tax credit. CAES is used to reduce the risk of committing uncertain quantities of wind energy and to shift dispatch of wind generation to high price periods. Using wind forecast data and market prices from 2006 – 2009, we find that the annual income for the modeled wind-CAES system would not cover annualized capital costs. We also estimate market prices with a carbon price of $20 and $50 per tonne CO2 and find that the revenue would still not cover the capital costs. The implied cost per tonne of avoided CO2 to make a wind-CAES profitable from trading on the day-ahead market is roughly $100, with large variability due to electric power prices. Wind power forecast errors for aggregated wind farms are often modeled with Gaussian distributions. However, data from several studies have shown this to be inaccurate. Further, the distribution of wind power forecast errors largely depends on the wind power forecast value. The few papers that account for this dependence bin the wind forecast data and fit parametric distributions to the actual wind power in each bin. A method to model wind power forecast uncertainty as a single closed-form solution using a logit transformation of historical wind power forecast and actual wind power data is presented in Chapter 3. Once transformed, the data become close to jointly normally distributed. We show the process of calculating confidence intervals for wind power forecast errors using the jointly normally distributed logit transformed data. This method has the advantage of fitting the entire dataset with five parameters while also providing the ability to make calculations conditioned on the value of the wind power forecast. The model present in Chapter 3 is applied in Chapter 4 to calculate increases in net load uncertainty introduced from day-ahead wind power forecasts. Our analyses uses data from two different electric grids in the U.S. having similar levels of installed wind capacity with large differences in wind and load forecast accuracy due to geographic characteristics. A probabilistic method to calculate the dispatchable generation capacity required to balance day-ahead wind and load forecast errors for a given level of reliability is presented. Using empirical data we show that the capacity requirements for 95% day-ahead reliability range from 2100 MW to 5600 MW for ERCOT and 1900 MW to 4500 MW for MISO, depending on the amount of wind and load forecast for the next day. We briefly discuss the additional requirements for higher reliability levels and the effect of correlated wind and load forecast errors. Additionally, we show that each MW of additional wind power capacity in ERCOT must be matched by a 0.30 MW day-ahead dispatchable generation capacity to cover 95% of day-ahead uncertainty. Due to the lower wind forecast uncertainty in MISO, the value drops to 0.13 MW of dispatchable capacity for each MW of additional wind capacity. Direct load control (DLC) has received a lot of attention lately as an enabler of wind power. One major benefit of DLC is the added flexibility it brings to the grid. Utilities in some parts of the U.S. can bid the load reduction from DLC into energy markets. Forecasts of the resource available for DLC assist in determining load reduction quantities to offer. In Chapter 5, we present a censored regression model to forecast load from residential air conditioners using historical load data, hour of the day, and ambient temperature. We tested the forecast model with hourly data from 467 air conditioners located in three different utilities. We used two months of data to train the model and then ran day-ahead forecasts over a six week period. Mean square errors ranged from 4% to 8% of mean air conditioner load. This method produced accurate forecasts with much lower data requirements than physics based forecast models.

Demand response and dynamic pricing are touted as ways to empower consumers, save consumers money, and capitalize on the “smart grid” and expensive advanced meter infrastructure. In this work, I attempt to show that demand response and dynamic pricing are more nuanced. Dynamic pricing is very appealing in theory but the reality of it is less clear. Customers do not always respond to prices. Price differentials are not always large enough for customers to save money. Quantifying energy that was not used is difficult. In chapter 2, I go into more detail on the potential benefits of demand response. I include a literature review of residential dynamic pilots and tariffs to see if there is evidence that consumers respond to dynamic rates, and assess the conditions that lead to a response. Chapter 3 explores equity issues with dynamic pricing. Flat rates have an inherent cross-subsidy built in because more peaky customers (who use proportionally more power when marginal price is high) and less peaky customers pay the same rates, regardless of the cost they impose on the system. A switch to dynamic pricing would remove this cross subsidy and have a significant distributional impact. I analyze this distributional impact under different levels of elasticity and capacity savings. Chapter 4 is an econometric analysis of the Commonwealth Edison RTP tariff. I show that it is extremely difficult to find the small signal of consumer response to price in all of the noise of everyday residential electricity usage. Chapter 5 looks at methods for forecasting, measuring, and verifying demand response in direct load control of air-conditioners. Forecasting is important for system planning. Measurement and verification are necessary to ensure that payments are fair. I have developed a new, censored regression based model for forecasting the available direct load control resource. This forecast can be used for measurement and verification to determine AC load in the counterfactual where DLC is not applied. This method is more accurate than the typical moving averages used by most ISO’s, and is simple, easy, and cheap to implement.

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.

This thesis addresses the cost-effectiveness of curtailing a wind farm to regulate the electrical grid frequency and the hurricane risk to offshore wind farms in the eastern United States. Additionally, this thesis presents a new method to generate long periods of non-stationary wind speed time series data sampled at high rates by combining measured and simulated data. Paper 1 calculates the cost of curtailing the power output of a wind farm to provide a reserve of power to regulate the electrical grid frequency, as required by grid operators in several countries with high wind-power penetrations. The simulations in Paper 1 show that it is most efficient to curtail a few turbines deeply rather than curtail all turbines in a wind farm equally. Compared to regulation prices in the Texas (ERCOT) market in 2007-2009, a curtailed wind farm would be cost-competitive with conventional generators less than 1% of the time. Paper 2 supports the simulations in Paper 1 by developing a method to combine long periods of low-frequency wind speed data with realistic simulated high-frequency turbulence. The combined time series of wind speeds retains the non-stationary characteristics of wind speed, such as diurnal variations, the passing of weather fronts, and seasonal variations, but gives a much higher sampling rate. Papers 3 and 4 estimate the hurricane risks to current designs of offshore wind turbines in the U.S. Paper 3 develops analytical probability distributions based on historical hurricane records to predict the distribution of damages to a single wind farm in a given location. Paper 4 uses simulated hurricanes with realistic statistical properties to estimate the correlated risks to all the wind farms in a region and estimate the distribution of aggregate losses over different periods. Both papers find hurricane risks are small for current turbine designs in New England and the Mid-Atlantic, but the risks in the Gulf of Mexico and the Southeast are significant enough to warrant new, stronger designs. Hurricane risks could be reduced almost an order of magnitude by ensuring that turbines can continue yawing to track the wind direction even if grid power is lost.

Content-centric networking (CCN) has emerged as a dominant paradigm for future Internet architecture design due to its efficient support for content dissemination, which currently dominates Internet use. This dissertation shows how economic and social welfare analysis can be used to inform the design of a CCN architecture that provides network stakeholders with incentives to deploy and use. Firstly, the dissertation investigates the economic incentives of different stakeholders to deploy content-centric network Internet architectures and shows that network operators will fail to deploy sufficient storage infrastructure to support CCN without payment ows from publishers. However, the level of payment required differs for different network players, which gives them different competitive advantages in providing storage infrastructure and content delivery services. Secondly, it evaluates the social welfare implications of different storage deployment scenarios in a CCN-based architecture and identifies two deployments that maximize social welfare. In the first, edge networks provide the storage infrastructure through a transaction broker. In the second, edge networks pay third-parties an amount, equivalent to the realized benefits from a storage node, to deploy storage infrastructure in the network. All other deployment scenarios lead to a deadweight loss. Thirdly, the dissertation identifies content delivery functionalities that break in a CCN-based architecture and shows how these functionalities can be successfully replicated and enhanced by a careful design of the structure of routable content, content naming and the meta-information added to content. The proposed design supports several content delivery applications and can be easily extended to other networking principals. Finally, the dissertation identifies and discusses threats in the CCN content delivery model and proposes some mechanisms to address these threats. In addition, the dissertation identifies some policy implications of the CCN content delivery model and proposes some policy interventions that may lead to desirable deployment outcomes.

The rapid growth of the shale gas industry has inspired questions concerning attendant apparent and potential short- and long-term health and environmental risks. My research examined three potential environmental and health risks. (1) For the last half-century the Northeast natural gas market was supplied from major producing areas in Texas, the Gulf Coast, and Canada. Because radon has a short half-life of 3.8 days, the time required to transport the natural gas from these areas to the Northeast resulted in a low-radon product being delivered to homes. As the Northeast gas market transitions to locally-produced natural gas the potential for radioactive decay will diminish and the natural gas being delivered to homes will contain radon at higher levels. I assess the lung cancer risk for people living in homes with unvented gas cooking (approximately half of the homes in the Northeast) and heating appliances, which are in fewer homes. Data on the locally-produced natural gas radon concentration are limited, but for the modeling assumptions considered the radon exposure is predicted to be small compared to typical residential exposures, and additional annual population-level risk will likely be much less than the error in the estimate of annual radon-induced lung cancers. An excess lifetime lung cancer risk >10-4 is possible for high gas usage in poorly ventilated settings. (2) High volume and locally-concentrated surface water withdrawals for Marcellus Shale development may pose a risk to water quality, aquatic and riparian ecosystems, and other uses of water resources. State environmental and interstate water authorities take different approaches to managing these water withdrawals. In the Upper Ohio River Basin, which covers the western third of Pennsylvania, the Department of Environmental Protection requires that all water used for shale gas development be covered by a water management plan. These plans stipulate the amount and timing of surface water withdrawals from each source as a function of annual stream flow statistics. Neighboring regulatory authorities and some environmental groups favor the use of monthly flow statistics instead, but implementation of these statistics in western Pennsylvania would require more data than are currently available. Because hydrologic data in the Upper Ohio River Basin are sparse, the use of the annual flow statistics is more likely than use of monthly flow statistics to prevent water withdrawals when aquatic ecosystems are under the greatest stress. The annual flow statistic might also result in fewer and smaller occurrences of computed ecodeficits under scenarios of development-related water demands in the future. (3) Improperly abandoned and orphan gas wells threaten human health and safety as well as pollute the air and water. Pennsylvania currently requires production companies to post a bond to ensure environmental reclamation of non-productive well sites, but the cost of plugging horizontally drilled wells and reclaiming well pads is estimated to be at least a factor of 10 greater than the current well bonds. The economics of shale gas development favor transfer of assets from large entities to smaller ones. With the assets go the liabilities, and without a mechanism to prevent the new owners from assuming reclamation liabilities beyond their means, the economics favor default on well-plugging and site restoration obligations. In addition to increasing the bond amounts, individual well trust accounts are proposed based on a model from the coal industry. Pre- and delayed-funding options (a fee and severance tax, respectively) to pay for future reclamation are examined from the perspective of the taxpayer. The exposure of the taxpayer to these financial liabilities and to a future orphan well problem can be minimized with minimal impacts to the profitability of gas production regardless of which funding option is used.

Battery cost is among the largest barriers to mainstream adoption of electric vehicles. This dissertation examines near future battery technology and cost by (1) validating existing physics-based battery performance models using laboratory testing and manufacturer specifications, (2) constructing battery design optimization and production cost models to identify the least-cost design and investigating how key design-decision variables affect performance and cost for a variety of vehicle power and energy requirements, and (3) conducting expert elicitation on future battery costs and the key factors that drive cost. The validation, cost, and optimization modeling work use LiNi0.33Co0.33Mn0.33O2/LixC6 (NMC-G) as the chemistry of choice. Validation results of Battery Design Studio™ (BDS) a Li-ion battery modeling software indicated that BDS predictions of total energy delivered under our constant C-rate battery discharge tests are within 6.5% of laboratory measurements for a full discharge and within 2.8% when a 60% state of charge window is considered. Once validated, BDS is used to develop a power meta-model that predicts the 10–sec power capability of a cell design as a function of its capacity (Ah) and cathode coating thickness (microns). The production cost model is a process-based model and is constructed adopting process step information from existing literature. Subsequently, an optimization model is developed which estimates the cheapest cost battery pack design for a set of five different electrified vehicles (EVs) whereby the role of design-decision variables like cathode coating thickness is investigated among others. The energy and power requirements for the EVs, used as constraints in the optimization model, are calculated using the Powertrain Systems Analysis Toolkit (PSAT). Battery pack costs calculated are in the range of costs reported in the literature. Results indicate that higher capacity cells manufactured using higher electrode coating thicknesses can decrease manufacturing costs by 5-8%. Results suggest that economies of scale can be reached at a plant size of about 200MWh. Expert elicitation indicates that a variation of NMC-G is likely to be the cheaper cell-chemistry by 2018 with no major technological breakthroughs. Some experts also expect manufacturing improvements resulting in higher electrode coating thicknesses and cell capacities expected by 2018.

Many studies have recently reported estimates of greenhouse gas (GHG) emissions and associated potential climate impacts of biofuel and natural gas (NG) use. U.S. corn ethanol production keeps increasing under federal mandates, and NG production soars due to successful tapping of unconventional resources in North America, particularly shale gas. Numerous life cycle assessment (LCA) studies document technology specific corn ethanol and NG GHG estimates. The estimates often include all life cycle stages from fuel supply to combustion, and point out potential for emissions reductions. Several studies suggest that using GHG emissions as an evaluation metric underestimates corn ethanol’s radiative forcing (RF) impact – a precursor and indicator for global temperature change – by 10-90% over the next few decades. This emissions timing effect may overestimate (i) ethanol’s climate benefits over gasoline and (ii) the effectiveness of U.S. policies mandating and subsidizing ethanol. This work revisits the above studies, and builds upon existing models to quantify RF impacts across the corn ethanol life cycle. The emissions timing factor (ETF) is significantly smaller than previous estimates (2-13% depending on the chosen impact time frame), and the effect is dwarfed by uncertainty in the GHG emissions estimates. Nevertheless, ETF reduces ethanol’s probability of meeting the federal target of 20% GHG reduction relative to gasoline from 53% (according to EPA GHG estimates) to 7-29%. However, the small potential climate impacts from U.S. ethanol use may not actually be observable based on estimated initial increases in global average surface temperature of < 0.001 °C. About 25% of global primary energy production comes from NG, whose life cycle GHG emissions and potential future climate impacts from substituting coal are highly uncertain due to fugitive methane (CH4) emissions from the NG industry. Accurately quantifying the NG fugitive emissions (FE) rate – the percentage of produced NG, mainly CH4 and ethane (C2H6) – released to the atmosphere is challenging due to the size and complexity of the NG industry. Recent LCA estimates suggest that the current NG FE rate could be as high as 8% and 6%, from shale and conventional NG, respectively, and other bottom-up studies indicate even higher rates several decades ago. This work analyzes possible ranges of the global average NG FE rate based on atmospheric CH4, C2H6, and carbon isotope (δ13C-CH4) measurements recorded since 1984, and top-down modeling of their sources and sinks. Box-model, δ13C-CH4mass balance, and 3D-modeling results agree on best estimate NG FE rates of 3-5% (of dry NG production and dry NG composition) globally over the past decade, and 5-8% around 1990. Upper bound FE rates are 5% and 7% in 2010 and 2000, respectively. Best estimate and upper bound values may be overestimated because both assume lower bound emissions from oil and coal production as well as complete absence of natural hydrocarbon seepage. While LCA studies are useful for identifying processes with the greatest NG FE reduction potential, the recent high bottom-up estimates do not appear representative of the U.S. national average based on top-down modeling results. Given the steadily declining NG FE rates one may expect that further emissions abatement is possible if industry practices are further improved.

This research considers a different way to assess autonomous ground robotic systems for implementation into society. For the past two decades, autonomous mobile ground robotic systems have existed in the science and technology domain, but so far, have had only limited applications in the field. An ongoing assumption is that robotic systems must be capable of fitting into an existing infrastructure and way of doing things—one defined for human operators. However, this assumption and the approach designed around it are not well-suited for the advancement of autonomous robotics, especially when complex operations are concerned. This can be attributed, in part, to the challenges in the object recognition and reasoning research areas. While recent strides have been made, the larger context of developing object recognition and reasoning technology to achieve a performance criteria developed for human operation will result in a less optimum path than if a new approach and subsequent integration path were taken. While clearly there are political and cultural constraints that limit the degree to which a new approach can be adopted, this work looks beyond these barriers, specifically considering autonomous robotics in military convoy and commercial snow clearing operations. This research uses a two-tiered approach that separately addresses critical utility-based decision criteria, focusing on technological viability and risk first and then cost and economic viability second. This approach, unlike a single-step net assessment, may be more reflective of the actual decision path. Additionally, by structuring the problem in this way, current risk abatement methods can be made more effective and new ones can be emplaced. Further, assumptions that are either inadvertently designed-in or that unfairly discount benefits and externalities associated with autonomous robotics can be re-visited. Thus, by adopting a new approach that fundamentally changes the underlying concept of operation, implementation of autonomous robotics into complex operations may be achieved much sooner or at substantially lower cost than if the current approach were maintained.