The behaviour of hydrocarbon fuels in contaminated groundwater systems is studied using a multicomponent reactive transport model. The simulated processes include residual NAPL dissolution, aerobic and anaerobic biodegradation with daughter-product transport, and transport of a reactive carrier with mixed equilibrium/kinetic sorption. The solution algorithm is based on a three-dimensional Galerkin finite element scheme with deformable brick elements and capacity for a free watertable search. Nonlinearities are handled through Picard iteration. Convergence is rapid for most applications and mass balance errors for all phases are minimal. The model is first applied to simulate a pilot scale diesel fuel dissolution experiment in which humic acid is used as a natural organic carrier to enhance dissolution and to promote biodegradation of the aqueous components. The pilot scale experiment is described by Lesage et al. (1995) and Van Stempvoort et al. (2000). The conceptual model includes 8 unique components dissolving from 500 mL of residual diesel fuel within a 3D saturated domain. Oxygen-limiting competitive aerobic biodegradation with a dynamic microbial population is also included. A mixed 2-site equilibrium/kinetic model for describing sorption of the carrier to the aquifer solids was adopted to reproduce the observed breakthrough of the humic acid and organic components. Most model parameters were obtained independently with minimal calibration. Batch sorption data were found to fit well at the pilot scale, however biodegradation and dissolution rates were not well known and had to be fitted. Simulations confirmed the observed 10-fold increase in effective solubility of trimethylnaphthalene, and increases on the order of 2-5 for methyl- and dimethylnaphthalene. The simulated plumes showed almost complete attenuation after 5 years, in excellent agreement with observed data. A sensitivity analysis showed the importance of carrier concentrations, binding coefficients, dissolution and biodegradation rates. Compared to a dissolution scenario assuming no carrier, the humic acid-enhanced dissolution case decreased the remediation time by a factor of about 5. The second application of the model involves simulating the effect of ethanol on the persistence of benzene in gasoline-impacted groundwater systems. The conceptual model includes a 4-component residual gasoline source which is dissolving at the watertable into a 3D aquifer. Comparisons are made between dissolved plumes from a gasoline spill and those from an otherwise equivalent gasohol spill. Simulations have shown that under some conditions, a 10% ethanol component in gasoline can extend the travel distance of a benzene plume by at least 150% relative to that from an equivalent ethanol-free gasoline spill. The increase is due to preferred consumption of oxygen by ethanol and a corresponding reduction in the biodegradation rate of benzene while the two plumes overlap. Because of differences in retardation however, the ethanol and benzene plumes gradually separate. The impact therefore becomes limited because oxygen rapidly disperses behind the ethanol plume and benzene degradation eventually resumes. A sensitivity analysis for two common spill scenarios showed that background oxygen concentrations, and benzene retardation had the most significant influence on benzene persistence. A continuous gasohol spill over 10 years was found to increase the benzene travel distance by over 120% and a pure ethanol spill into an existing gasoline plume increased benzene travel distance by 150% after 40 years. The results are highly relevant in light of the forthcoming ban of MTBE in California and its likely replacement by ethanol by the end of 2002.

Ethanol can affect the biodegradation of gasoline hydrocarbons in groundwater. High concentrations of ethanol can be toxic to subsurface microorganisms that are otherwise capable of degrading hydrocarbons, such as benzene, toluene, ethylbenzene and xylenes (BTEX). At lower concentrations, ethanol may hinder BTEX degradation through substrate competition and the depletion of inorganic nutrients (e.g., nitrogen and phosphorus), oxygen and other electron acceptors needed for BTEX degradation. A series of laboratory experiments were designed to study the effect of ethanol on aquifer microorganisms and on aerobic BTEX biodegradation. A microcosm experiment was conducted to investigate the effect of ethanol on the biodegradation of BTEX. Microcosms were set up with Borden aquifer material and groundwater in which oxygen and nutrients were not limited. These microcosms contained BTEX in combination with a range of ethanol concentrations. Under these favourable conditions, the presence of ethanol up to concentrations of 1.9% (v/v) (equivalent to 15000 mg/L) caused little inhibition of BTEX degradation. Further experiments were conducted to study the antimicrobial effects of higher concentrations of ethanol. Following exposure to ethanol concentrations of 25% (v/v) or higher, microbial activity and survival was significantly diminished. Results suggest that a high concentration ethanol slug will have a major impact on the microbial community but that there would likely be potential for recovery. The recovery potential was examined further in laboratory column experiments designed to simulate a dynamic field situation where a high ethanol pulse is followed by a BTEX plume. These column experiments were conducted with Borden aquifer material and groundwater under aerobic conditions. The concentration of the ethanol pulse was 25% (v/v), which was expected to significantly alter the microbial population without destroying it. Following the ethanol exposure, groundwater and BTEX were allowed to flow through one column to simulate the reinoculation of microorganisms from upgradient groundwater advecting into the contaminated zone. The other column was fed with sterile groundwater and BTEX to evaluate the regeneration of within-column microorganisms that survived the ethanol exposure. Recovery in both columns was rapid. Unfortunately, during the recovery phase, sterility of the influent groundwater could not be maintained. As a result, recovery by regeneration could not be evaluated. Nonetheless, it is evident that recovery in terms of aerobic BTEX biodegradation was significant under the conditions of the column experiment. Ethanol did not appear to pose a long-term impact on BTEX biodegradation when oxygen and nutrients were in excess. In field situations, nutrients and electron acceptors may be limited; however, ethanol toxicity is not likely to cause a prolonged inhibition of BTEX biodegradation.

Brownfields have an important impact on health. They can influence physical health by increasing risk for health hazards such as the potential for injury hazards, disease transmission, or exposure to chemicals. They can also influence social health determinants like neighborhood level social capital or behavioral risk factors. Reusing brownfields for community gardens reduces environmental hazards and associated health hazards. It further promotes public health, and sustainable quality environment. Community gardens increase nutrition access, especially for many in low income populations, and community aesthetic. They also strengthen social cohesion and create recreational or therapeutic opportunities for a community, becoming part of the urban green space network. Special care must be taken to protect public health when reusing a brownfield for a community garden, like sampling for chemicals, cleaning up soil, and using protective garden designs. The overall benefit to the community is worth the initial investment required.

Because of the impact on human health and sustainable livelihood, the topic of drinking water and sanitation facilities is becoming a seriously discussed issue among international organizations as well as developing agencies in industrialized countries. The importance of water and sanitation management initialized this master thesis.   The aim of the project is to do an assessment of the drinking water quality as well as the sanitation situation in the village Condrita in the Republic of Moldova. This was done by studying the existing water and sanitation facilities, sampling the water, evaluating the reason for the poor water quality and mapping the current situation using ArcGIS. Furthermore, technologies for improving the drinking water and sanitation facilities are suggested.   The work was carried out by doing a literature study on how water sources and sanitation facilities should be constructed in order to ensure people’s health and to meet their needs. Geographic coordinates and water samples were collected from twenty-two public wells and springs. Interviews on the water situation and sanitation facilities were performed. Furthermore, water samples were analysed with respect to nitrate, turbidity, electrical conductivity and coliform bacteria amongst others. Pesticide contamination was also taken into consideration when one of the wells was analysed. Water sources were classified as improved or unimproved according to definitions by WHOSIS. Moreover, the DRASTIC vulnerability model was used to evaluate the groundwater susceptibility to contaminants.   In general, the water quality in the study area was poor and measured values of the analyzed parameters exceeded international or Moldovan standards for nitrate, hardness, electrical conductivity and total coliform bacteria. Four wells were contaminated with E. coli bacteria. Furthermore, turbidity measurements exceeded Moldovan standards in seven out of twenty-two water sources. No pesticide contamination was detected. Sampled water from the densely populated parts of the village as well as unimproved water sources proved to be of poorer quality. Map results showed that a majority of the groundwater within the study area was subject to a moderate or high risk of becoming contaminated. The current sanitation situation is that most families use simple pit latrines, which are placed far away from the dwelling-houses. Digging a new toilet when the existing one is full is a common practice in Condrita. Hand-washing facilities are seldom placed in proximity to the toilets.   Pit latrines are believed to be the most important source of groundwater contamination in the study area. Other sources are agricultural activities and poor practice when abstracting water from the wells. A feasible solution to improve both the drinking water quality and the sanitation situation would be to install ecosan toilets. Improvements of the well’s features that are suggested include construction of an apron slab as well as proper lids for covering the well.

The behaviour of hydrocarbon fuels in contaminated groundwater systems is studied using a multicomponent reactive transport model. The simulated processes include residual NAPL dissolution, aerobic and anaerobic biodegradation with daughter-product transport, and transport of a reactive carrier with mixed equilibrium/kinetic sorption. The solution algorithm is based on a three-dimensional Galerkin finite element scheme with deformable brick elements and capacity for a free watertable search. Nonlinearities are handled through Picard iteration. Convergence is rapid for most applications and mass balance errors for all phases are minimal. The model is first applied to simulate a pilot scale diesel fuel dissolution experiment in which humic acid is used as a natural organic carrier to enhance dissolution and to promote biodegradation of the aqueous components. The pilot scale experiment is described by Lesage et al. (1995) and Van Stempvoort et al. (2000). The conceptual model includes 8 unique components dissolving from 500 mL of residual diesel fuel within a 3D saturated domain. Oxygen-limiting competitive aerobic biodegradation with a dynamic microbial population is also included. A mixed 2-site equilibrium/kinetic model for describing sorption of the carrier to the aquifer solids was adopted to reproduce the observed breakthrough of the humic acid and organic components. Most model parameters were obtained independently with minimal calibration. Batch sorption data were found to fit well at the pilot scale, however biodegradation and dissolution rates were not well known and had to be fitted. Simulations confirmed the observed 10-fold increase in effective solubility of trimethylnaphthalene, and increases on the order of 2-5 for methyl- and dimethylnaphthalene. The simulated plumes showed almost complete attenuation after 5 years, in excellent agreement with observed data. A sensitivity analysis showed the importance of carrier concentrations, binding coefficients, dissolution and biodegradation rates. Compared to a dissolution scenario assuming no carrier, the humic acid-enhanced dissolution case decreased the remediation time by a factor of about 5. The second application of the model involves simulating the effect of ethanol on the persistence of benzene in gasoline-impacted groundwater systems. The conceptual model includes a 4-component residual gasoline source which is dissolving at the watertable into a 3D aquifer. Comparisons are made between dissolved plumes from a gasoline spill and those from an otherwise equivalent gasohol spill. Simulations have shown that under some conditions, a 10% ethanol component in gasoline can extend the travel distance of a benzene plume by at least 150% relative to that from an equivalent ethanol-free gasoline spill. The increase is due to preferred consumption of oxygen by ethanol and a corresponding reduction in the biodegradation rate of benzene while the two plumes overlap. Because of differences in retardation however, the ethanol and benzene plumes gradually separate. The impact therefore becomes limited because oxygen rapidly disperses behind the ethanol plume and benzene degradation eventually resumes. A sensitivity analysis for two common spill scenarios showed that background oxygen concentrations, and benzene retardation had the most significant influence on benzene persistence. A continuous gasohol spill over 10 years was found to increase the benzene travel distance by over 120% and a pure ethanol spill into an existing gasoline plume increased benzene travel distance by 150% after 40 years. The results are highly relevant in light of the forthcoming ban of MTBE in California and its likely replacement by ethanol by the end of 2002.

Ethanol can affect the biodegradation of gasoline hydrocarbons in groundwater. High concentrations of ethanol can be toxic to subsurface microorganisms that are otherwise capable of degrading hydrocarbons, such as benzene, toluene, ethylbenzene and xylenes (BTEX). At lower concentrations, ethanol may hinder BTEX degradation through substrate competition and the depletion of inorganic nutrients (e.g., nitrogen and phosphorus), oxygen and other electron acceptors needed for BTEX degradation. A series of laboratory experiments were designed to study the effect of ethanol on aquifer microorganisms and on aerobic BTEX biodegradation. A microcosm experiment was conducted to investigate the effect of ethanol on the biodegradation of BTEX. Microcosms were set up with Borden aquifer material and groundwater in which oxygen and nutrients were not limited. These microcosms contained BTEX in combination with a range of ethanol concentrations. Under these favourable conditions, the presence of ethanol up to concentrations of 1.9% (v/v) (equivalent to 15000 mg/L) caused little inhibition of BTEX degradation. Further experiments were conducted to study the antimicrobial effects of higher concentrations of ethanol. Following exposure to ethanol concentrations of 25% (v/v) or higher, microbial activity and survival was significantly diminished. Results suggest that a high concentration ethanol slug will have a major impact on the microbial community but that there would likely be potential for recovery. The recovery potential was examined further in laboratory column experiments designed to simulate a dynamic field situation where a high ethanol pulse is followed by a BTEX plume. These column experiments were conducted with Borden aquifer material and groundwater under aerobic conditions. The concentration of the ethanol pulse was 25% (v/v), which was expected to significantly alter the microbial population without destroying it. Following the ethanol exposure, groundwater and BTEX were allowed to flow through one column to simulate the reinoculation of microorganisms from upgradient groundwater advecting into the contaminated zone. The other column was fed with sterile groundwater and BTEX to evaluate the regeneration of within-column microorganisms that survived the ethanol exposure. Recovery in both columns was rapid. Unfortunately, during the recovery phase, sterility of the influent groundwater could not be maintained. As a result, recovery by regeneration could not be evaluated. Nonetheless, it is evident that recovery in terms of aerobic BTEX biodegradation was significant under the conditions of the column experiment. Ethanol did not appear to pose a long-term impact on BTEX biodegradation when oxygen and nutrients were in excess. In field situations, nutrients and electron acceptors may be limited; however, ethanol toxicity is not likely to cause a prolonged inhibition of BTEX biodegradation.

Ingestion exposure estimates typically use a default bioavailability of 100%, thereby assuming that the entirety of an ingested dose is absorbed into systemic circulation. However, the actual bioavailability of ingested contaminants is oftentimes lower than 100%. The research described herein investigates the use of the Simulator of the Human Intestinal Microbial Ecosystem (SHIME) for the calculation of <i>in vitro</i> bioaccessibility (IVBA), a conservative predictor of bioavailability, of mercury (Hg) from traditional northern foods and arsenic (As) from soils. The primary objective of the research described herein is to address data-gaps which have hindered attempts to incorporate IVBA into risk assessment on more than a case-by-case basis. The hypotheses of this thesis are that (1) the bioaccessibility of contaminants is dependent upon concentration due to kinetic limitations on dissolution, (2) gastrointestinal (GI) microbes in the ileum and colon alter contaminant bioaccessibility and/or speciation, (3) the GI microbial effect on bioaccessibility is toxicologically relevant, and (4) metal bioaccessibility is predictable according to dissolution kinetics.<p> Mercury bioaccessibility from country food samples was independent of total Hg concentration (F=0.5726, P=0.578) whereas As bioaccessibility was inversely related to total As concentration for Nova Scotia mine tailings, synthesized ferrihydrite with adsorbed AsV, and synthesized amorphous scorodite (P=2 x 10-10). Isotherm analysis indicated that, at high soil As concentrations, saturation of simulated GI fluids limited As bioaccessibility under gastric conditions whereas kinetic limitations constrained As bioaccessibility under intestinal conditions. Additionally, we demonstrated that GI microbes may affect Hg bioaccessibility, either increasing or decreasing bioaccessibility depending upon the type of food. For example, the bioaccessibility of HgT decreased in the presence of GI microbial activity for caribou kidney, caribou tongue, seal blood, seal brain, seal liver, and walrus flesh. In contrast, HgT bioaccessibility from Arctic char and seal intestine increased in the presence of GI microbial activity. Similarly, colon microbial activity increased (Fishers Protected LSD, P<0.05) As bioaccessibility from synthesized amorphous scorodite (56 110%), Nova Scotia mine tailings (140 300%), an agricultural soil (53%) and an ironstone soil (350%) containing elevated arsenic concentrations. However, under small intestinal conditions, this microbial effect was transient and demonstrated a small effect size. The toxicological relevance of microbial effects upon As bioaccessibility was assessed using a juvenile swine model with co-administration of oral antibiotics (neomycin and metronidazole). This study research indicated that microbial effects on As bioaccessibility are not reflected in the juvenile swine model. For example, the microbial communities present in the pigs proximal colon clustered according to antibiotic treatment (e.g. microbial communities of antibiotic treated pigs differed from non-treated pigs). Despite this, the urinary arsenic excretion (and hence arsenic bioavailability) of antibiotic-treated juvenile swine orally exposed to soil-borne arsenic was equivalent (Holm-Sidak, P=0.930) to the urinary arsenic excretion of juvenile swine not treated with antibiotics. Therefore, in vitro GI models may not need to include a microbially active intestinal stage when measuring As IVBA.<p> Metal bioaccessibility from soils appears predictable according to fundamental chemical properties of the metal-of-concern. Specifically, metal bioaccessibility of 7 of the 13 metals (V, Ni, Zn, Cu, U, Cd, & Ba but not Tl, Pb, As, Se, Cr, and Hg) regulated according to Canadian Council of Ministers of the Environment Soil Quality Guidelines (CCME SQG) were strongly dependent (R2 = 0.7) on water exchange rate constants of metal cations (kH20) indicating that desorption kinetics may serve as the foundation of a predictive model of metal bioaccessibility.

The Swedish industrial society has left several areas with contaminated soil and water which today is of danger for the environment and human health. One of these areas with contaminated soil and water is the region where the former Glava glassworks operated during 1859 to 1939. The main object in this study has been to study the responsibility for the contaminations according to the Environmental Code. The method used is a qualitative content analysis of relevant literature to answer the following questions: What types of contaminations can be found in the area of Glava glasswork and what health issues can these cause? What kind of activity has been conducted in the area that have caused the contaminations in soil and water? According to law, what are the responsibilities for the contaminations in soil and water at Glava glassworks? The contamination in the region has been confirmed to originate from the glasswork due to the environmental engineering survey made in the area in 2009. The contaminations found in the area are arsenic, zinc, boron and other heavy metals which can cause severe damage in the environment and people’s health. A historical analysis has been crucial to study the situation of responsibility. The time dimension is crucial according to Swedish environmental law, the Environmental Code (SFS 1998:808). To find the responsible of the contamination in the region, the contaminations must have occurred after the 30 June 1969 according to 8 § the law (SFS 1998:811) on the introduction of the Environmental Code or someone must have obtained the property after the 1 January 1999 according to 15 § the same law (SFS 1998:811). None of these criteria have been fulfilled which means that the Swedish state is responsible of investigation and remediation of the contaminations in the area. / Det svenska industrisamhället har efterlämnat flera områden med föroreningar i mark och vatten som idag är en fara för både människors hälsa och miljön. Ett av dessa är området kring Glava glasbruk som bedrev sin verksamhet mellan 1859 och 1939. I denna kvalitativa studie, där en kvalitativ innehållsanalys av relevant litteratur har använts, är syftet att undersöka hur ansvarsförhållandena ser ut för föroreningarna på glasbruksområdet där följande frågeställningar har legat som grund. Vilka är föroreningarna på Glava glasbrukområdet och vilka hälsoskador kan dessa ge upphov till? Vilka verksamheter har bedrivits på platsen som kan ha lett till uppkomsten av föroreningarna? Hur ser ansvarsförhållandena ut enligt miljöbalken för mark- och vattenföroreningarna vid Glava glasbruk? Föroreningar på platsen har konstaterats härstamma från bruket i och med en genomförd miljöteknisk markundersökning som gjorts på området under 2009. De föroreningar som hittats på platsen är bland annat arsenik, zink, bor och andra tungemetaller som kan ge allvarliga skador människors hälsa och på miljön. För att undersöka ansvarsförhållandena har en historisk analys av platsen gjorts. En djupdykning i miljöbalken har visat att tidsaspekten har varit avgörande. För att en ansvarig ska kunna ställas till svars måste föroreningarna ha uppkommit efter den 30 juni 1969 enligt 8 § Lag (SFS 1998:811) om införande av miljöbalken eller fastigheten måste ha förvärvats efter den 1 januari 1999 enligt 15 § samma lag (SFS 1998:811). Ingen av dessa kriterier uppfylls här vilket då innebär att ansvaret för undersökning och efterbehandling ligger hos staten.

Soil and groundwater at many existing and former industrial areas and disposal sites is contaminated by organic solvent compounds that were released into the environment. Organic solvent compounds are heavier than water. When they are released into the subsurface, they tend to adsorb onto the soils and cause the appearance of LNAPL (light nonaqueous phase liquid) and DNAPL (dense nonaqueous phase liquid) pool. The industrial petroleum hydrocarbons (e.g., methyl tertiary-butyl ether, MTBE and benzene) and chlorinated solvent (e.g., trichloroethylene, TCE) are among the most ubiquitous organic compounds found in subsurface contaminated environment. One cost-effective approach for the remediation of the chlorinated solvent and petroleum products contaminated aquifers is the installation of permeable reactive zones or barriers within aquifers. As contaminated groundwater moves through the emplaced reactive zones, the contaminants are removed, and uncontaminated groundwater emerges from the downgradient side of the reactive zones. The objectives of this study were developed to evaluate the feasibility of applying in-situ chemical oxidation (ISCO) barrier and in-situ slow polycolloid-releasing substrate (SPRS) biobarrier system on the control of petroleum hydrocarbons and chlorinated solvent plume in aquifer. In the ISCO barrier system, it contained oxidant-releasing materials, to release oxidants (e.g., persulfate) contacting with water for oxidating contaminants existed in groundwater. In this study, laboratory-scale fill-and-draw experiments were conducted to determine the compositions ratios of the oxidant-releasing materials and evaluate the persulfate release rates. Results indicate that the average persulfate-releasing rate of 7.26 mg S2O82-/d/g was obtained when the mass ratio of sodium persulfate/cement/sand/water was 1/1.4/0.24/0.7. The column study was conducted to evaluate the efficiency of in situ application of the developed ISCO barrier system on MTBE and benzene oxidation. Results from the column study indicate that approximately 86-92% of MTBE and 95-99% of benzene could be removed during the early persulfate-releasing stage (before 48 pore volumes of groundwater pumping). The removal efficiencies for MTBE and benzene dropped to approximately 40-56% and 85-93%, respectively, during the latter part of the releasing period due to the decreased persulfate-releasing rate. Results reveal that acetone, byproduct of MTBE, was observed and then further oxidized completely. Results suggest that the addition of ferrous ion would activate the persulfate oxidation. However, excess ferrous ion would compete with organic contaminants for persulfate, causing the decrease in contaminant oxidation rates. In the SPRS biobarrier system, the food preparation industry has tremendous experiences in producing stable oil-in-water (W/O, 50/50) emulsions with a uniformly small droplet size. Surfactant mixture (71 mg/L of SL and 72 /L of SG) blending with water could yield a stable and the optimal emulsion was considered the best. The small absolute value of the emulsion zeta potential reduces inter-particle repulsion, causing the emulsion droplets to stick to each other when they collided. Overtime, large masses of flocculated droplets can form which then clog the sediment pores. The results can be used to predict abiotic interactions and distribution of contaminant mass expected after SPRS injection, and thus provides a more accurate estimate of the mass of TCE removed due to enhanced biodegradation. The effect of TCE partitioning to the vegetable oil on contaminant migration rates can be approximated using a retardation factor approach, where 0.28 years through a 3 m barrier. In anaerobic microcosm experiments, result show that SPRS can be fermented to hydrogen and acetate could be used as a substrate to simulate reductive dehalorination. The apparent complete removal of nitrate and sulfate by SPRS addition was likely a major factor that promoted the complete reduction of TCE at later stages of this study. Results from the column experiment indicate that occurrence of anaerobic reductive dechlorination in the biobarrier system can be verified by: (1) the oil: water partition coefficients of dissolved TCE into vegetable oil were be used to predict abiotic interactions and distribution of contaminant mass expected after SPRS injection. (2) The SPRS can ferment to hydrogen and acetate could be used as a substrate to simulate reductive dechlorination. The proposed treatment scheme would be expected to provide a more cost-effective alternative to remediate other petroleum hydrocarbons and chlorinated solvents-contaminated aquifers. Experiments and operational parameters obtained from this study provide an example to design a passive barriers system for in-site remediation.

Windback seals work similarly to labyrinth seals except for the effect of helical groove. These seals are essentially a tooth on stator or tooth on rotor labyrinth seal where the grooves are a continuous helical cut like a thread. Windback seals are used in centrifugal gas compressor to keep oil out of the gas face seal area. These face seals cannot be contaminated by oil. A purge gas is applied to the seal to help force the oil back into the bearing area. The windback seal should be designed to prevent any oil contamination into the supply plenum and to reduce purge gas leakage. The CFD simulations have been performed with the effect of clearance, tooth width, cavity shape, shaft rotation, eccentricity, and tooth location on the seal leakage performance and the flow field inside the seal. The leakage flow rate increases with increasing the pressure differential, rotor speed, radial clearance, cavity size, and shaft diameter and with decreasing the tooth width. The eccentricity has a minimal effect for the windback seal. From oil simulations, the windback seal with 25% rotor eccentricity has some of the journal bearing action and drives back flow into the gas plenum. However the windback seal can be used to force the oil back into the bearing side before starting the compressor by applying a purge gas flow since the positive axial velocity inside the cavity is larger than the negative axial velocity. m A Rw cav & / ? is constant for varying shaft rotation since the leakage flow rate for the windback seal increases linearly as the the rotor speed increases. The leakage flow rate for the windback seal increases as the groove size increases due to the pumping action of the windback seal. A windback seal design based upon the numerical simulations that minimize gas leakage and help prevent gas face seal oil contamination was optimized. The windback seal has two leakage flow paths. Since the leakage flow rate under teeth of windback seals is the same as for a similar geometry labyrinth seal, the flow under the teeth can be predicted by two-dimensional labyrinth seal analysis. An empirical model for the leakage rate through the cavity has been developed which fits the data with a standard deviation of 0.12.