A rise of 2 oC in the Earth’s temperature is likely to occur when the concentration of CO2 in the atmosphere reaches approximately 450 ppm. CO2 emissions are closely related to the continual use of fossil fuels. In order to make fossil fuels sustainable, carbon capture & storage (CCS) is required to reduce CO2 emissions. There are three leading CO2 capture methods, namely post-combustion capture, oxy-fuel combustion and integrated gasification combined cycle (IGCC) pre-combustion capture. CO2 hydrate (CO2:6H2O) formation has been investigated as a way to capture CO2 in the IGCC conditions. The formation of hydrate in this work was experimentally investigated in an isochoric system (batch mode) inside a vertical fixed bed reactor (FBR), also known as high pressure volumetric analyser (HPVA). Standard silica gel with an average particle size of 200-500 µm, mean pore size of 5.14 nm, a pore volume of 0.64 cm3/g and a surface area of 499 m2/g was used as a porous medium. The presence of hydrate in FBR was justified by using graphic methods. The solubility of CO2 in water using Henry’s Law and the experimental pressure–time (P-t) curve were analysed to determine the formation of hydrate. Hydrate formation was confirmed when the mole fraction of CO2 dissolved in water exceeded the Henry’s Law value as well as a two-stage pressure drop in the experimental P-t curve. Initially, various sample preparation methods (methods 1, 2, 3 and 4) were studied leading to the selection of method 4 (the use of vigorous stirring) which had the highest moisture content (14.8 wt%) and the greatest water conversion to hydrate (40.5 mol%) at 275 K and 36 bar in a pure CO2 gas system. Also, high regeneration and repeatability of the results for all samples prepared by method 4 were expected as less water was occluded inside silica gel pores. Further investigations in pure CO2 gas systems highlighted the effect of type of silicas used, the importance of the type of promoters used, the concentration of promoters, experimental driving force, silica pore size, bed height and the amount of moisture content for formation of hydrate. Standard silica gel was the only silica found to show hydrate formation due to the best distribution of pore size. The high amount of bulk water inside zeolites 13X and spherical MCF-17 (21.3 and 50.8 wt% respectively) was the main reason of no hydrate formation observed. Additionally, the combined-promoters designated type T1-5 (0.01 mol% sodium dodecyl sulphate (SDS) + 5.6 mol% tetrahydrofuran (THF)) and type T3-2 (0.01 mol% SDS + 0.1 mol% tetra-butyl ammonium bromide (TBAB)) were the two best obtaining a CO2 uptake of 5.95 and 5.57 mmol of CO2 per g of H2O respectively. Ethylene glycol mono-ethyl ether (EGME; 0.1 mol%) was a good alternative to THF when combined with SDS (0.01 mol%) with a CO2 uptake of 5.45 mmol of CO2 per g of H2O for this combined-promoter designated type T1A-2. In addition, the CO2 uptake increased as ∆P increased or ∆T decreased. Moreover, mesoporous silica (silica gel) performed better than microporous silica (zeolite 13X) where the formation of hydrate by zeolite 13X was observed with minimal CO2 uptake (0.58 mmol of CO2 per g of H2O) when the bed height was reduced. Additionally, the total amount of CO2 consumed through hydrate formation increased as the amount of water inside mesoporous silica increased which was not the case for microporous silica. Furthermore, the experiments performed in the IGCC conditions (283 K and 70 bar) by employing T1-5 and T3-2 in a fuel gas mixture demonstrated low hydrate formation with a CO2 uptake of 1.5 and 1.1 mmol of CO2 per g of H2O respectively. This was expected due to the slow kinetics since CO2 molecules were competing with H2 molecules which also reduced the selectivity of CO2 molecules during hydrate formation. Hence, in reality, pure CO2 system is the best option for CCS through hydrate formation at the right operating conditions as compared to fuel gas mixture.

Metal pollutants such as copper released into the aqueous environment have been increasing as a result of anthropogenic activities, a topic causing global concern. Adsorption-based treatment technologies offer opportunities to remediate metal pollutants from municipal and industrial wastewater effluent. The aim of this work was to evaluate the capability of modified cellulose nanowhisker (CNW) adsorbents for the remediation of copper from water matrices under realistic conditions using response surface methodology (RSM) and artificial neural network (ANN) models. The first part of the study explored the preparation and characterisation of modified CNW adsorbents. It also focused on the stability of the modified CNW adsorbents at different time intervals under dry conditions (up to 28 days) and in the water matrix (up to 7 days). The results showed that the modified CNW adsorbents were stable at different time intervals under dry conditions and in the water matrix and proved that the functional groups were permanent and did not degrade under the tested conditions. The stability of these modified CNW adsorbents under these conditions, which is relevant from both the manufacturing and application perspectives, is reported for the first time in this study. The second part of the work focused on using copper as a case study for heavy metal pollution in a clean water matrix, to evaluate removal by modified CNWs under several conditions and ranges appropriate to wastewater treatment plants (WWTPs), using factorial experimental design. RSM and ANN models were employed in order to optimise the system and to create a predictive model to evaluate the Cu(II) removal performance by the modified CNW adsorbents. Moreover, unseen experiments not belonging to the training data set, located both inside and outside the test parameter system, were performed to test the model suitability. This is also novel, as generally only one or two parameter variations have been tested, without checking the chosen model suitability for parameters lying between the tested parameters, and certainly not for parameters lying outside the tested parameter space, as has been done in this study. The results obtained showed that the ANN model outperformed the RSM model when predicting copper removal from a clean water matrix. The Langmuir andFreundlich isotherm models were applied to the equilibrium data, and the results revealed that the Langmuir isotherm (R2 = 0.9998) had better correlation than the Freundlich isotherm (R2 = 0.9461). Experimental data was also tested in terms of kinetics studies using pseudo-first order and pseudo-second order kinetic models. The results showed that the pseudo-second-order model accurately described the kinetics of adsorption. The third part of the work was aimed at gaining a deeper understanding of the complexity and variability of the wastewater matrix, including evaluating the impact of the wastewater matrix temporally on adsorbent performance to remediate copper pollutant from a real-world wastewater matrix. This study has demonstrated that the wastewater matrix composition, which is both complex and variable, has an impact on adsorbent capability and performance. A benchmark study was adopted as a ‘new’ water quality parameter to inform on the effects of the wastewater matrix (wastewater composition and its variability) on the modified CNW adsorbent’s capability to remediate copper from this matrix. Since the process of adsorption from wastewater is often complicated due to the variation in wastewater composition, results obtained from the benchmark experiments were included as one of the independent variables in ANN modelling, unlike in other optimisation studies. The performance of the ANN and RSM models was statistically evaluated in terms of coefficient of determination (R2), absolute average deviation (AAD), and root mean squared error (RMSE) on predicted experimental outcomes. The ANN model including the variability of wastewater composition fitted the experimental data with excellent accuracy and better prediction (R2 = 0.9963) than both the ANN model that did not include this variability (R2 = 0.9945), and the RSM model (R2 = 0.9409). The outcome of this study showed that by supplying the ANN model with the data obtained from the benchmark experiments as the fourth independent variable, it was possible to improve the predictability of the ANN model. Continuous flow experiments for remediation of spiked Cu(II) from the wastewater matrix were conducted. However, the physical structure of modified CNW adsorbents renders them unsuitable for use in column operation. Therefore, a more detailed study of the mechanical properties of CNW adsorbents would be necessary in order to improve the strength and stability of the adsorbents. This work has demonstrated that modified CNW are promising adsorbents to remediate copper from water matrices under realistic conditions including wastewater complexity and variability. The use of models to predict the test parameter system and account for matrix variability when evaluating CNW adsorbents for remediating Cu from a real-world wastewater matrix may also provide the foundation for assessing other treatment technologies in the future.

Membrane filtration and Advanced Oxidative Processes (AOPs) are among the most efficient and cost-effective methods employed in water purification. A system to integrate the two methods using photoactive colloidal particles was studied in this thesis, with the final purpose of overcoming membrane fouling, one of the main issues occurring in filtration processes. The production of nanostructured TiO2 microparticles through a simple and extremely rapid synthesis and an easy method to assemble a multifunctional coating, integrating inorganic particles on filtration membranes, were targeted as the most promising solutions from the technological and environmental point of view. The control of microwave-assisted heating applied to hydrothermal treatments, a relatively recent synthetic method, allowed the production of nanostructured mesoporous spherical TiO2 particles, bringing the synthesis to the minute scale, extremely rapid compared with conventional heating, and achieving products otherwise difficult to obtain without the help of surfactants or templating agent. The as-synthesised particles showed photoactivity under visible light, with rate of specific reactions (selective de-ethylation) 4 times higher compared with commercial photocatalysts. Furthermore, the particles were modified to extend the limited intrinsic absorbance of TiO2 in the visible light, with promising results given by formation of stoichiometric defects (in particular oxygen vacancies) through annealing under vacuum. This treatment allowed the achievement of comparable or even higher performance in photocatalytic degradation of rhodamine B with respect to commercial TiO2 photocatalysts, including Aeroxide P25, with degradation rate towards organic molecules (rhodamine B) of even 60-70% after 1 hours, compared to the 25% of P25. The production of a multifunctional coating for water treatment by integration of colloidal and nanometric TiO2 particles has been also studied. A simple technique to integrate TiO2 nanoparticles onto different substrate, in particular filtration membranes, was developed by simple electrostatic interactions involving the use of polyelectrolytes, water-soluble charged polymer forming organised layers when assembled in a macromolecular structure defined as polyelectrolyte multilayers (PEMs). Electrostatic assembly was applied as an environmentally friendly technique to anchor nanoparticles (P25) on different surfaces, transferring their properties to these. In particular, the application of TiO2 particles conferred hydrophilic and superhydrophilic to a relatively hydrophobic surface (Mylar) by controlling the multilayer assembly conditions, in particular the ionic strength of the polyelectrolyte solutions. The achievement of superhydrophilic behaviour on the treated surfaces, with contact angles below 15° on Mylar surfaces, and the possibility of removing fouled active layer from a membrane replacing it with a newly generated one can be both implemented as potential antifouling strategies in water treatment.

The hydrodynamic vortex separator (HDVS) is currently employed at wastewater treatment works and in the sewerage system as a combined sewer overflow (CSO) for the separation of solids from an incoming waste stream. This project presents the first stage in developing and aiding the existing design methodology for the optimisation of kinetic processes within the HDVS. The kinetic process design methodology combines hydraulic and kinetic principles by using the true mixing regime characteristics of a system and batch reactor data to determine a kinetic processes efficiency. This project used residence time distribution (RTD) analysis to extensively characterise the mixing regime within a model and prototype HDVS. The HDVS was operated with and without a baseflow component and with and without the sludge hopper for a range of inlet flow rates and flow splits covering design flow rates for a number of existing applications. The RTD was obtained using a pulse tracer injection method and described using the complete range of data analysis techniques typical employed in RTD studies. This includes the axial dispersion model (ADM), tanks-in-series model (TISM), RTD indices and a RTD combined mathematical model. The combined model is configured to quantify the inactive flow behaviour within the HDVS i. e. stagnant and dead volumes. The HDVS has a complex imperfect plug-flow mixing regime. This non-ideal flow behaviour is associated with both dispersion and dead volumes and results in short-circuiting. At low flow rates the HDVS operating without a baseflow contains fluid elements which conduct flow slower than the mean velocity. At high flow rates the inactive flow behaviour is associated with dead volumes and subsequently short-circuiting. The flow rate at which this change in mixing characteristics occurs is termed the transition flow rate and is approximately 151/min and 901/min for the model and prototype HDVS respectively. At all flow rates above the transition flow rate the HDVS has a very stable mixing regime, which is associated with both the inactive flow behaviour and the plug-flow mixing characteristics. The ADM and TISM parameters increase as the flow rate decreases and therefore, the HDVS has improved plug-flow mixing characteristics and reduced dispersion at low flow rates. Removing the sludge hopper reduces the inactive flow behaviour and improves the plug-flow mixing characteristics. The inactive flow behaviour within the model HDVS operating with no baseflow occupies approximately 20-40% of the total volume and similarly for the prototype HDVS 5-25% and increases as the inlet flow rate increases. The inactive flow behaviour occupies a smaller fraction of the total volume and the plug-flow mixing characteristics are also improved as the HDVS is scaled-up in size. Hence, the scale-up of the HDVS will provide a mixing regime with less short-circuiting and improved plug-flow mixing characteristics and therefore, more conducive for certain kinetic processes and particularly chemical disinfection processes. The introduction of a baseflow component alters the total mixing regime within the HDVS. The baseflow component introduces an element of plug-flow mixing and subsequently the total plug-flow mixing characteristics of the HDVS operating with a baseflow component are greater than the HDVS operating without a baseflow. The baseflow component plug-flow mixing characteristics increase and the overflowcomponent decrease as the inlet flow rate increases. Short-circuiting of the baseflow and overflow component occurs as the inlet flow rate decreasesa nd increasesr espectively. Hence, there are different mixing regimes within the HDVS associated with the overflow and baseflow component. The HDVS operating with a baseflow component has improved plug-flow mixing characteristics when the sludge hopper is included. This project was also extended to include an experimental kinetic process analysis, by investigating the first-order decomposition of hydrogen peroxide (H202) using catalase. This was undertaken to compare the actual kinetic process performance within the HDVS to that estimated using the RTD. The H202 decomposition results showed that the design of the HDVS for kinetic processes can be achieved using only the RTD and relevant batch reactor data. This enables the HDVS to be optimised for kinetic process applications and eliminates the need for costly and time consuming pilot trials. The characterisation of the HDVS using RTD analysis creates scope for significant future research. This includes: alternative experimental RTD techniques, development of the RTD combined mathematical model to include a baseflow component and kinetic process principles, extensive kinetic process batch reactor investigations, application of both the hydraulic and kinetic data into chemical reactor design computer software and finally the scaling of the HDVS using the RTD and therefore the kinetic process optimisation. This work is a proactive response by practitioners and Hydro International Plc to pressure from the regulators and EU Directives, placing emphasis on the use of sophisticated treatment processes based on good scientific principles, to meet current and future stringent water quality standards.

In the UK, the Water Services Regulation Authority (OFWAT), estimates that for all the water and sewage companies there was over 3,365 Ml/d (megalitres per day) of water leakage in the reported period 2010 and 2011, which was still at 3094.21 Ml/day in the 2013 review. Leakage estimates do not include water siphoned illegally through unaccounted connections worsened as asset management of buried utility services has been overlooked for years. With the asset management programme framework (AMP6) due to come into effect in April 2015, water companies are expected to get more out of their existing infrastructure and making considerable investment while keeping bills reasonable for customers. Improving the management of these assets is therefore a priority, as effective asset management enables companies to reduce cost, through leakage management, to plan investments and repairs, and to evaluate operational risks by better fault prediction rather than the current reactive approach. This research focuses on the water distribution network as an asset that include leaks and pipe infrastructure with different materials and diameters. A novel method for leak detection and asset management using an electromagnetic sensor has been developed. Trials in the laboratory showed the sensor is capable of detecting pipe types and conditions thus improving leak detection and asset management in the water industry without extensive digs and modification to existing access valves within the network for system deployment. Furthermore, the sensor would potentially benefit the gas industry by modifying the frequency of operation to match gas filled cylindrical metal structures.

This research aims to propose an integrated method, which combines all the aspects required to reduce environmental impact from waste packaging and to increase knowledge on the best way to enhance reusable packaging. Through a review of the extant literature, a conceptual framework was designed of the most important dimensions to enhance reusable packaging amongst society and industries The main contributions in the research are the development of a Social Behaviour Aspect Model (SBAM) and the creation of reusable packaging attributes checklist. The SBAM can help industries focus on having high knowledge about reuse of packaging and to cooperate with communities to develop personal and social values and norms during the designing of reusable packaging. SBAM is the output from the first phase, which showed the importance of making an effort to develop packaging for consumers to reuse. The reusable packaging attributes checklist can provide a guideline for manufacturers/designers who intend to develop packaging sustainability performance through designing reusable packaging, and contribute to meet and interpret the reuse of packaging requirements and procedures. It also determines the environmental impact of reusable packaging attributes, which many industries are concerned about. The reusable packaging attributes checklist is the output from the second and third phases. The System Dynamic (SD) method was the approach used to determinate the interaction between social aspects and reusable packaging. The Normal Average, Codes and Coding and factor analysis with Principal Component Analysis (PCA) approaches were used to determine the reusable packaging attributes checklist. The last phase of the research was the conduction of a case study of a real company which needs to increase the amount of reusable packaging it uses and which seeks to reduce its environmental impact. All methods used in this research have both a quantitative and a qualitative nature. Data was collected by evaluation of consumers' responses and experts' experiences, as provided in the questionnaires. This research opens up opportunities for improving packaging and meeting sustainable profits and provides valuable information based on social, economic and environmental benefits of reusable packaging. The novelty of this research can help industries to investigate the most important areas for development within communities to enhance the use of reusable packaging and also facilitate the process-based change from one-way packaging to reusable packaging effectively with reduction of environmental impact.

The efficacy of pond treatment systems is dependent on the internal hydrodynamic and mixing interactions between aquatic vegetation and the adjacent flow. In attempting to improve pollution degradation and reduce the effects of hydraulic short circuiting, an understanding and quantification of these interactions was sought for seasonal changes in vegetation growth. Controlled laboratory studies were conducted using detailed Laser Induced Fluorometry (LIF) and Ultrasound Velocity Profiling (UVP) techniques to quantify mixing across vegetated shear layer, emergent Cattail reeds (Typha latifolia). An Optimised Finite Difference Model (OFDM) was developed to predict the best fit downstream concentration distributions given the input profiles of transverse mixing coefficient, Dy(y). The model provided strong fitting in artificial vegetation (R2 = 0.977 and 0.969 for high and low density rigid cylinders). A good fitting was also made for the winter reeds (R2 = 0.976); although the physical application of conventional shear layer theory failed to significantly improve predictions in the summer season reeds above those of a simple discontinuity functionality describing Dy(y). The form of the lateral variation in transverse mixing coefficient was confirmed in the artificial vegetation studies where peak mixing is enhanced by shear layer vortices. However, in real vegetated shear flows, the heterogeneities in stem morphology and distribution render the relative magnitude of shear layer mixing diminished when compared to other regions of the flow. It is shown that, while the OFDM provides good predictions of concentration distributions when using a physically justified profile of the transverse mixing coefficient, a discrete step formulation is sufficient for describing mixing in real vegetated shear flows. This study shows therefore, that, while shear layer mixing is dominant in artificial, uniform vegetation, transverse mixing in real vegetated flows is dominated by complex geometries, localised shear processes and bed roughness effects.

The Nigerian economy has relied heavily on crude oil production since independence in 1960. As a consequence, it has seen an influx of multinational petroleum companies with oil exploration and associated activities having significant environmental impacts, particularly oil leakage and spillage into soil and the overall degradation of the ecosystem in the Niger Delta area. This study aims to find a viable solution to the remediation of polluted soil by comparing two thermal remediation techniques, namely microwave pyrolysis and traditional pyrolysis, which has been investigated using a Gray-King retort. The polluted soil was first examined to ascertain the distribution of the soil organic carbon (SOC) with 78% found to be solvent extractable in dichloromethane/methanol, while 95 % was thermally labile and removed under hydropyrolysis (HyPy) conditions at 550 °C. The remaining 5 % of the SOC was composed of a recalcitrant residue being defined as the black or stable polyaromatic carbon fraction. The solvent extractable organic matter (EOM) was then further separated into the maltene (free phase) and asphaltene (bound phase) fractions together for comparison with a sample of Nigerian crude oil provided by the Shell Petroleum Development Company (SPDC), Nigeria. The Nigerian crude oil is a light crude oil with the percentage of maltene (95.2 %) was far higher than the asphaltene (4.8 %). A closer margin was observed in the percentage between the maltene (88.3 %) and asphaltene (11.7 %) in the soil EOM due to biodegradation. The biomarker profile of the EOM was compared with that of a Nigerian crude oil to confirm that the EOM contains the crude oil in the soil. Their biomarker profiles revealed that the source inputs were terrigenous from deltaic settings, of Late Upper Cretaceous age and deposited under oxic conditions. Oleanane (a pentacyclic triterpene, abundant in oils from the Niger Delta) was present in both the crude oil and EOM and the hopane and the sterane distributions (m/z 191 and m/z 217 respectively) were similar in every respect, which indicates that the probable source of the pollutant crude oil in the soil is similar in composition to the Nigerian crude oil. Accordingly, the polluted soil was treated with microwave pyrolysis and Gray-King pyrolysis to remove the crude oil pollutant. The maximum average recovered products from the thermal remediation process with Gray-King pyrolysis is 99.4 % TOC and maximum crude oil pollutant removed by Gray-King pyrolysis was 85.3 % TOC with maximum oil recovery of 70 % TOC from all the different treatment conditions, while the shortest treatment time condition gave the lowest gas yield of 10.2 % TOC. This implies that 100 % removal with respect to EOM and 89 % removal with respect to HyPy as discussed above. Furthermore, the polluted soil was also treated with microwave pyrolysis with maximum pollutant removal of 77 % TOC, which is 98.7 % removal with respect to EOM and 81 % with respect to HyPy. In conclusion, Gray-King pyrolysis removed more of the soil organic carbon than microwave pyrolysis, but the latter does have advantages regarding operability and greater output within a short treatment time.

The primary aims of this research was firstly to develop a computer modelling tool which could predict pollution retention in a rain garden and secondly to use the model and additional experiments to examine various aspects of rain garden design with respect to pollutant retention. Initially, the behaviour of all contaminants in urban runoff was examined including their retention and possible modelling methods. Heavy metals were then identified as the main focus of this project as this choice was the most beneficial addition to current research. The main factors affecting their retention were found to be macropore flow, pore water velocity, soil moisture content and soil characteristics and the primary method of modelling capture was identified as a sorption isotherm. Thus a dual-permeability heavy metal sorption model was developed; this was based on an intensive literature review of current best practice in both hydrological modelling and pollutant retention fields with respect to rain garden devices. The kinematic wave equation was chosen to model water movement in both the matrix and macropore regions as this provided a simpler alternative to more complex equations while still maintaining good accuracy. With regards to the modelling of heavy metal retention three isotherms were chosen: the linear, Langmuir and Freundlich equations as these were found from previous research to be the most accurate. These isotherms were incorporated into a one dimensional advection-dispersion-adsorption equation in order to model both transport and retention together. This model was tested against the appropriate literature and accurate comparisons were obtained thus validating it. Column experiments were designed to both provide a unique contribution to rain garden research and further validate the model. This was achieved by analysing past experiments and identifying an area where research is lacking; this area was the effect of macropore flow on heavy metal retention in rain garden systems under typical English climatic conditions. The findings of these experiments indicated that although macropore flow did not impact the hydraulic performance of the columns, retention of the most mobile of heavy metals, copper, was decreased slightly in one case. The overall retention of the columns was still high however at a value in excess of 99% for copper, lead and zinc. The results of the experiments were also used to further validate the model. The model was then applied to the development of a rain garden device for a planned roundabout in Kent, U.K. Preliminary design considered an upper root zone layer with organic soil and a sandy storage sublayer each 30 cm thick, for a rain garden area of 5 and 10% the size of the contributing impervious surface. Two scenarios were examined; the accumulation and movement of metals without macropores and the possibility of groundwater contamination due to preferential flow. It was shown that levels of lead can build up in the upper layers of the system, but only constituted a health hazard after 10 years. Simulations showed that copper was successfully retained (no significant concentrations below 50 cm of rain garden soil depth). Finally given concerns of preferential flow bypassing sustainable drainage systems, macropore flow was examined; results indicated that due to site conditions it was not a threat to groundwater at this location for the time frame considered. These actions successfully completed the objectives of this project and it was deemed successful.

Dispersion in sewers affects the concentration of pollutants within urban drainage systems and the quest to understand it has led to extensive studies of the hydraulics within circular manholes. Since there are both circular and square manholes within the urban drainage system, it is important to understand the hydraulics within a square manhole, its similarities and/or differences. This research is aimed at describing the transportation and mixing of soluble pollutants through urban drainage systems and the effects of structures such as manholes. It also seeks to study the relationship between headloss and dispersion coefficient in a full flowing pipe. Laboratory solute transport studies were carried out on a straight pipe with two diameter sizes (35 mm and 25 mm) of orifice plate inserted. Two square manhole sizes (388 mm x 388 mm and 150 mm x 150 mm) with a straight through flow was also studied at surcharge and overflow conditions. A relationship between the total headloss along the length of the pipe and the dispersion coefficient was established. With the introduction of a new parameter, a relationship was also established between the headloss due to the orifice and the dispersion coefficient when an orifice plate is inserted along the line of flow. It was observed for the large manhole size that at low surcharge depths there was more mixing occurring while at high surcharge depths the dye travels in a straight jet with some tracer trapped at the upper manhole volume. There exists a transition zone between the low and high surcharge depths which is not evident in the small size manhole. These observations mirror the phenomena that have already been observed in circular manholes. Manhole overflow studies was indicative of a completely mixed flow for the large manhole whereas for the small manhole it suggests a piston flow (sometimes referred to as plug flow) with some longitudinal dispersion occurring.