Doctor of Philosophy / Department of Biological & Agricultural Engineering / Stacy L. Hutchinson and Kyle R. Douglas-Mankin / Ecohydrology, a sub-discipline of hydrology, deals with the ecological impacts of and interactions with the hydrological cycle. Changes in hydrology of the Great Plains rivers, and their impacts on water quality, water resources, aquatic ecosystems, and fish species distributions have been documented. The major goal of this study was to develop and test methods to analyze watershed-level ecohydrological characteristics. The specific objectives were (a) to detect past temporal trends and spatial variability in hydrologic indices, (b) to evaluate the presence and/or extent of spatial and temporal relationships between climatic and ecohydrological variables and riverine historical data on fauna species density and distribution, and (c) to assess model calibration strategies for accurate ecohydrological indicator simulation. The Kansa River Basin (KRB), which has substantial land use, soil and climate variability, as well as variation in anthropogenic drivers (dams, diversions, reservoirs, etc.), was the focus of this study. Thirty eight hydrological indicators were generated using the indicators of hydrologic alterations software for 34 stations in the KRB using 50-year streamflow records and trend analysis using Mann-Kendall, Seasonal Kendall, and Sen’s slope estimator tests. Across the KRB a decreasing trend was evident for annual mean runoff, summer and autumn mean runoff, 30-day, 90-day minimum flows, and 1-day, 3-day, 7-day, 30-day and 90-day maximum flows. Most of the significant negative trends were observed in the High Plains ecoregion. Two hydrologic indicators, high-flow pulse count and mean summer streamflow, were significantly different in streams that lost two indicator fish species, indicating that changes in streamflow have altered the fish habitat of this region. The Soil and Water Assessment Tool (SWAT) biophysical model calibrated using a multi-objective framework (multi-site, multivariable and multi-criteria) was able to simulate most of the ecohydrological indicators at different hydrological conditions and scales. The SWAT model provided robust performance in simulating high-flow-rate ecohydrologic indicators. However ecohydrologic indicators performance was highly dependent on the level of calibration and parameterization. The effect of calibration and parameterization on ecohydrologic indicators performance varied between watersheds and among subwatersheds.

Master of Science / Department of Biological and Agricultural Engineering / Stacy L. Hutchinson / Coal-fired generation accounts for 45% of the United States electricity and generates harmful emissions, such as sulfur dioxide. With the implementation of Flue Gas Desulfurization (FGD) systems, sulfur dioxide is removed as an air pollutant and becomes a water pollutant. Basic physical/chemical wastewater treatment can be used to treat FGD wastewater, but increased regulations of effluent water quality have created a need for better, more economical wastewater treatment systems, such as constructed wetlands. At Jeffrey Energy Center, north of St. Mary’s, KS, a pilot-scale constructed wetland treatment system (CWTS) was implemented to treat FGD wastewater before releasing the effluent into the Kansas River. The objectives of this study were to 1.) determine if a portable water quality meter could be used to assess water quality and track pollutant concentrations, 2.) develop a water balance of the CTWS, 3.) generate a water use coefficient for the CWTS, and 4.) create a mass balance on the pollutants of concern. Water quality measurements were taken with a HORIBA U-50 Series Multi Water Quality Checker and compared to analytical water tests provided by Continental Analytic Services, Inc. (CAS) (Salina, KS). The water balance was created by comparing inflows and outflows of data determined through flow meters and a Vantage Pro2™ weather station. Information from the on-site weather station was also used to compute the system water use coefficient. Water sampling was conducted from date to date at 10 locations within the CWTS. In general, there was little to no relationship between the HORIBA water quality measurements and the analytical water tests. Therefore, it was recommended that JEC continue to send water samples on a regular basis to an analytical testing laboratory to assess the CWTS function and track pollutants of concern. Because the water balance was conducted during system initiation, there was a great deal of fluctuation due to problems with the pumping system, issues with the upstream FGD treatment system, extreme weather events, and immature vegetation. This fluctuation resulted in the system having a non-steady state operation, which weakened the ability to calculate a system water use coefficient. However, during periods of strong system function, the water use coefficient was similar to previous studies with maximum water use being approximately equal to the reference evapotranspiration. The results of the mass balance indicated high removals mercury, selenium, and fluoride, but low removals of boron, manganese, chloride, and sulfate were exported from the CWTS.

Master of Science / Department of Chemical Engineering / James Edgar / In the past decade the technique of horizontal drilling and hydraulic fracturing has been improved so much that it has become a cost effective method to extract natural gas from shale formations deep below the earth’s surface. Natural gas extraction has boomed in the past few years in the United States, enough that it has driven prices to an all time low. The amount of natural gas reserves in the U.S. has led to claims that it can lead the country to energy independence. It has also been touted as a cleaner fuel for electricity generation and to power vehicles. This report explains hydraulic fracturing and horizontal drilling particularly with regards to utilizing the techniques for natural gas extraction from shale gas. It also discusses the environmental impact due to the drilling and gas extraction. It demonstrates that although the natural gas beneath the U.S. is a valuable resource, the impacts to the planet and mankind are not to be taken lightly. There is the potential for the effects to be long term and detrimental if measures are not taken now to control them. In addition although on the surface natural gas seems to be a greener fuel, particularly in comparison to gasoline, it is also considered worse for the environment.

Master of Science / Department of Biological and Agricultural Engineering / Kyle R. Douglas-Mankin / Aleksey Y. Sheshukov / According the U.S. Environmental Protection Agency’s (EPA) website, our planet is at risk of global warming due to greenhouse gas emissions. The earth’s average temperature has been reported to have risen by 1.4°F over the last century. This seemingly small increase in average planetary temperature has been linked to devastating floods, severe heat waves, and dangerous and unpredictable shifts in our climate (US EPA, 2013a). In the 2012 report, the Intergovernmental Panel on Climate Change states that bioenergy has the potential to significantly mitigate greenhouse gases as long as this is produced in a sustainable manner (Chum, et al., 2011). In light of these facts, research into the sustainable production of bioenergy sources in the United States is currently underway. To ensure that the correct biofuel crop is selected for a given region and to investigate any secondary effects of changing our nation’s agricultural practices to include biofuels, computer models can be very useful. The Soil Water Assessment Tool (SWAT) is a robust, continuous time step model that was developed by the USDA Agricultural Resource Service that can simulate changes in land use and land management and the effect this has on erosion, water quality, and other important factors. This paper describes the preliminary work to create a model of the Kanopolis Lake Watershed that is part of the Kansas River Basin using SWAT 2009. Data pertaining to weather, topography, land use, management, stream flow, and reservoirs was gathered and incorporated into the SWAT model. This was then simulated to obtain the uncalibrated data. SWAT produced unacceptable statistics for both crop yields and for stream flow using the Nash-Sutcliffe Efficiency equation and using percent bias. This suggests that the model must be calibrated to be of use in understanding both the current and future land use scenarios. Once the model is calibrated and validated, it can be used to simulate different biofuel cropping scenarios.

Master of Science / Department of Civil Engineering / Alok Bhandari / In the upper Midwest, subsurface tile drainage water is a major contributor of nitrate (NO[subscript]3–N) coming from fertilizers and animal manure. Movement of NO[subscript]3-N through tile drainage into streams is a major concern as it can cause eutrophication and hypoxia conditions, as in the Gulf of Mexico. Denitrifying bioreactors is one of the pollution control strategies to treat contaminated tile drainage water. These bioreactors require four conditions which are: 1) organic carbon source, 2) anaerobic conditions, 3) denitrifying bacteria and 4) influent NO[subscript]3-N. This research focuses on investigating fate of veterinary antibiotics in woodchips commonly used in in-situ reactors. Tylosin (TYL) and sulfamethazine (SMZ) are two veterinary antibiotics which are most commonly used in the United States and can be found in tile water after manure is land applied. Partition coefficients of TYL and SMZ on wood were determined by sorption experiments using fresh woodchips and woodchips from an in situ reactor. It was concluded that the woodchips were an effective means to sorb the veterinary antibiotics leached into the tile water after application of animal manure. Linear partition coefficients were calculated and phase distribution relationships were established for both the chemicals. The fresh woodchips gave inconclusive data but predictions could be made by the information determined in the experiments using woodchips from a ten year old woodchip bioreactor. Desorption was also studied and the likelihood of desorption was predicted using the Apparent Hysteresis Index. Overall, it was found that the old woodchips allowed for quick sorption of both antibiotics. It was also found that SMZ had reversible sorption on old woodchips. Thus, it was concluded that the woodchip bioreactor would not be effective for removal of veterinary antibiotics from tile drainage. More research is required for the fate of TYL and to confirm the conclusion.

Master of Science / Department of Biological and Agricultural Engineering / Naiqian Zhang / A real time wireless, optical sensor network was tested for long-term, remote monitoring of suspended sediment concentrations (SSC) in streams. The sensor and control board assembly was calibrated using a two-stage calibration procedure, including a pre-calibration conducted in the laboratory to adjust the sensitivity of the sensor and a field calibration using grab samples to establish an effective statistical model to predict SSC from the sensor signals. The assembly was installed in three military bases around the United States. These bases were Fort Riley, Kansas; Fort Benning, GA; and Aberdeen Proving Ground, MD. The types of water bodies and watersheds varied greatly among the sites, which allowed the sensor to be tested under versatile conditions for potential widespread use. The results show that the sensor was capable of measuring SSC at each watershed independently. The calibration model developed for each sensor can be used to predict SSC from real-time sensor data. A data processing algorithm was developed to lessen the effect of fouling and clogging on sensor signals, along with eliminating anomalies in the data gathered. The results of this study displayed meaningful prediction data that can be used to estimate SSC in a stream over a long period of time. Information obtained in this study can be used as a launching point for future work and understanding of stream processes.

Master of Science / Biological & Agricultural Engineering / Stacy L. Hutchinson / Increasing attentiveness to climate change and the dependence of human life on natural resources has spurred awareness about the detrimental impacts of human activity on the environment. Ecosystem services, or the benefits that humans derive from ecosystems, have changed more in the past 50 years than in any other comparable period in human history (Carpenter et al., 2009).The dilemma of managing the trade-off between immediate human needs and maintaining the ability of the Earth to provide ecosystem services is considered to be one of the largest challenges of this century (Foley et al., 2005). The ecosystem service concept aims maximize the provision of services across an entire ecosystem to achieve overall ecosystem health through land management, policy, and economic decisions. The intent of this research was to improve such decisions by increasing the understanding about the relationship between urban best management practices and freshwater provision, erosion regulation, and flood regulation ecosystem services. Fifty-six land management scenarios with varying densities of BMP application were simulated using the Stormwater Management Model (SWMM). The ecosystem services resulting from these land management scenarios were quantified using indices developed by Logsdon and Chaubey (2013). Results demonstrate that the application of bioretention cells improve both freshwater provision and erosion regulation services immediately downstream from the implementation site, and an increase in erosion regulation services was observed at the greater watershed scale. There was no change in the provision of freshwater, erosion regulation, or flood regulation services observed by the application of green roofs or rain barrels at either scale of analysis.

Doctor of Philosophy / Department of Biological and Agricultural Engineering / Ronaldo G. Maghirang / Emission of particulate matter (PM) and various gases from open-lot beef cattle feedlots is becoming a concern because of the adverse effects on human health and the environment; however, scientific information on feedlot emissions is limited. This research was conducted to estimate emission rates of PM[subscript]10 from large cattle feedlots. Specific objectives were to: (1) determine feedlot PM[subscript]10 emission rates by reverse dispersion modeling using AERMOD; (2) compare AERMOD and WindTrax in terms of their predicted concentrations and back-calculated PM[subscript]10 emission rates; (3) examine the sensitivity of both AERMOD and WindTrax to changes in meteorological parameters, source location, and receptor location; (4) determine feedlot PM[subscript]10 emission rates using the flux-gradient technique; and (5) compare AERMOD and computational fluid dynamics (CFD) in simulating particulate dispersion from an area source. PM[subscript]10 emission rates from two cattle feedlots in Kansas were determined by reverse dispersion modeling with AERMOD using PM[subscript]10 concentration and meteorological measurements over a 2-yr period. PM[subscript]10 emission rates for these feedlots varied seasonally, with overall medians of 1.60 and 1.10 g /m[superscript]2 -day. Warm and prolonged dry periods had significantly higher PM emissions compared to cold periods. Results also showed that the PM[subscript]10 emissions had a diurnal trend; highest PM[subscript]10 emission rates were observed during the afternoon and early evening periods. Using particulate concentration and meteorological measurements from a third cattle feedlot, PM[subscript]10 emission rates were back-calculated with AERMOD and WindTrax. Higher PM[subscript]10 emission rates were calculated by AERMOD, but their resulting PM[subscript]10 emission rates were highly linear (R[superscript]2 > 0.88). As such, development of conversion factors between these two models is feasible. AERMOD and WindTrax were also compared based on their sensitivity to changes in meteorological parameters and source locations. In general, AERMOD calculated lower concentrations than WindTrax; however, the two models responded similarly to changes in wind speed, surface roughness, atmospheric stability, and source and receptor locations. The flux-gradient technique also estimated PM[subscript]10 emission rates at the third cattle feedlot. Analyses of PM[subscript]10 emission rates and meteorological parameters indicated that PM[subscript]10 emissions at the feedlot were influenced by friction velocity, sensible heat flux, temperature, and surface roughness. Based on pen surface water content measurements, a water content of at least 20% (wet basis) significantly lowered PM[subscript]10 emissions at the feedlot. The dispersion of particulate from a simulated feedlot pen was predicted using CFD turbulence model ([kappa]-[epsilon] model) and AERMOD. Compared to CFD, AERMOD responded differently to wind speed setting, and was not able to provide detailed vertical concentration profiles such that the vertical concentration gradients at the first few meters from the ground were negligible. This demonstrates some limitations of AERMOD in simulating dispersion for area sources such as cattle feedlots and suggests the need to further evaluate its performance for area source modeling.

Master of Science / Department of Mechanical and Nuclear Engineering / Steven J. Eckels / Membrane dehumidification and evaporative cooling applications have the potential to significantly improve the energy efficiency of air conditioning equipment. The use of wire mesh media in such membrane applications is feasible but has not been studied extensively. Therefore, the aim of this work is to investigate the heat and mass transfer performance of several different wire mesh media in membrane based dehumidification and evaporative cooling. There were six wire mesh membranes tested in an experimental facility. The wire mesh membranes vary with respect to percent open area, wire diameter, pore size and material. Two non-permeable, solid membranes were also tested in the facility and compared with the wire mesh membranes. The test section of the experimental facility consists of a narrow air duct and a plate apparatus. The membrane samples were fashioned into rectangular plates and installed into the test section. The plate membranes separate liquid water and air flow streams. The inlet air temperature and humidity are altered to produce condensation or evaporation at the membrane surface. The average convective heat and mass transfer coefficient of the air boundary layer is measured for each of the experimental plates. Membrane based dehumidification and evaporative cooling were accomplished using the wire mesh media. However, the wire mesh membranes did not exhibit any significant differences in their performance. The mesh plates were compared with the solid plate membranes and it was discovered that the solid plates exhibited significantly higher heat transfer coefficients during condensation conditions. This result most likely is due to the formation of large water droplets on the solid plates during condensation. The experimental data is then compared to analytical predictions of the heat and mass transfer coefficients developed from several heat transfer correlations and by invoking the heat and mass transfer analogy. The experimental data is also compared directly with the heat and mass transfer analogy. It was found that the data did not compare well with the heat and mass transfer analogy. This result is attributed to the fact that the membrane surface limits the amount of direct exposure to the gas-liquid interface.

Doctor of Philosophy / Department of Biological and Agricultural Engineering / Ronaldo Maghirang / Emission of greenhouse gases (GHGs), including nitrous oxide (N2O), methane (CH4), and carbon dioxide (CO2), from open beef cattle feedlots is becoming an environmental concern; however, scientific information on emissions and abatement measures for feedlots is limited. This research was conducted to quantify GHG emissions from feedlots and evaluate abatement measures for mitigating emissions. Specific objectives were to: (1) measure N2O emissions from the pens in a commercial cattle feedlot; (2) evaluate the effectiveness of surface amendments in mitigating GHG emissions from feedlot manure; (3) evaluate the effects of water application on GHG emissions from feedlot manure; and (4) compare the photo-acoustic infrared multi-gas analyzer (PIMA) and gas chromatograph (GC) in measuring concentrations of N2O and CO2 emitted from feedlot manure. Field measurements on a commercial beef cattle feedlot using static flux chambers combined with GC indicated that N2O emission fluxes varied significantly with pen surface condition. The moist/muddy surface had the largest median emission flux; the dry and compacted, dry and loose, and flooded surfaces had significantly lower median emission fluxes. Pen surface amendments (i.e., organic residues, biochar, and activated carbon) were applied on feedlot manure samples in glass containers and evaluated for their effectiveness in mitigating GHG emissions. Emission fluxes were measured with the PIMA. For dry manure, all amendments showed significant reduction in N2O and CO2 emission fluxes compared with the control (i.e., no amendment). For moist manure, biochar significantly reduced GHG emissions at days 10 and 15 after application; the other amendments had limited effects on GHG emissions. The effect of water application on GHG emissions from feedlot manure was evaluated. Manure samples (with and without water application) were placed in glass containers and analyzed for GHG emission using a PIMA. For the dry manure, GHG emissions were negligible. Application of water on the manure samples resulted in short-term peaks of GHG emissions a few minutes after water application. Comparison of the GC and PIMA showed that they were significantly correlated but differed in measured concentrations of N2O and CO2. The PIMA showed generally lower N2O concentrations and higher CO2 concentrations than the GC.