No description available.

Early approaches in engineering safety were based on deterministic concepts, according to which, safety was assured by making conservative assumptions in the design of industrial systems, as well as through the use of safety factors, based on engineering judgement. Recent safety approaches are based on probabilistic concepts, in reliability techniques and accident consequence modelling, according to which safety is defined in terms of probabilities or frequency of failures and severity of consequences. These two items compose the basis for the evaluation of industrial risks. Risk assessment techniques are particularly relevant for those working offshore, who are exposed full-time to several different hazards from process equipment, risers or blow-outs, ship collisions, helicopter accidents, extreme weather, etc. At present, there are several sophisticated techniques, models and software already developed, that can be used in probabilistic risk analysis. However, general criteria for the acceptability of estimated risks have not been suitably developed, although several attempts have been made, and the question: "How safe is safe enough?" still remains without an appropriate answer. Within this context, this Thesis presents a brief description of techniques and models presently used for risk evaluation, as well as legislation concerning risk tolerability criteria and other proposed safety targets. Regarding the establishment of global safety criteria for offshore oil units, the focus of this work will be the development of a risk and reliability allocation methodology to achieve them. From data collected from more than thirty Brazilian offshore oil units, some of them operating for more then ten years, individual risk values for offshore employees are going to be calculated. Based on these values, obtained for the individual risk, and on a quantitative risk assessment performed for a Brazilian Floating Production Storage Offloading vessel, a risk and re1iability allocation methodology will be proposed. This methodology provides a feasible" model to allocate reliability and risk criteria for the main safety functions of an offshore unit in a self-consistent manner. It provides a method for design engineers to establish minimum reliability levels for the safety systems of an industrial facility, in order to achieve safety targets previously defined for it. The application of a risk and reliability allocation model to the problem of setting criteria for a range of hazardous scenarios presented in the operation of an offshore oil production unit is a novel approach. It is hope that the proposed methodology can contribute to a wider discussion about the establishment of a global measure for the evaluation of safety performance of offshore oil units.

A combination of rising oil demand and declining supply from the conventional sources is drawing global attention to the vast heavy-oil resources. These are commonly developed with steam-based processes which, in most cases, burn fossil fuel to generate the required steam. However, tightening constraints on fuel, water, and the environment are some of the factors currently fuelling the interests in enhancements to the traditional steaming operations. To mitigate some of the steam-related issues, we introduce two new thermal recovery methods, namely: (i) alternating-injection of steam and CO2 (SAC), and (ii) alternating-injection of steam and flue-gas (SAF). The primary objective of this research is to assess the technical and commercial feasibility of these new processes. To achieve this objective, we employ a combination of analytic modelling, numerical simulations and experimental studies, investigating the reservoir heat-transport aspects of steam-based processes, asphaltene-induced formation impairment, as well as the key controls on reservoir dynamics and project economics. In this work, the concepts of first-contact condensation (FCC) and multiple-contact condensation (MCC) have been introduced as additional mechanisms of heat-transport in steam-based processes. Hence, the traditional conductive-convective heat equations have been extended. Solutions of these equations indicate that laboratory and field observations are better rationalised, hence eliminating the current practice of employing unrealistic effective permeability and thermal diffusivity to explain these observations. We also provide conditions under which petroleum reservoirs may be analysed as adiabatic systems, and establish the relative influence of reservoir and operating parameters on reservoir heat-transport. Considering the asphaltene-precipitation potentials of CO2 and flue-gas, new models have been formulated for describing asphaltene-induced impairment of the permeability of porous media which, in turn, have been analysed as either closed (non-flowing) or open (flowing) systems. Application of the models to rationalise the experimental results from common porous media, which include sandstone, carbonate and glass-bead, validates their robustness. As a further test on the robustness of the proposed models, their main underlying assumptions have been validated with a set of capillary-flow experiments, which approximate asphaltene deposition at pore scale. As a case study for reservoir simulations, the Nigerian heavy-oil deposit has been examined. The sensitivity of reservoir response to reservoir, geometric (number and design of wells) and operating parameters has been quantified. From these results, a realistic set of dynamic-simulation models has been constructed for the Nigerian deposit. Within the parameter-space explored, the main subsurface uncertainties are reservoir geometry, permeability distribution as well as fluid and relative-permeability models. In addition, all the processes, namely steam-alone, SAC and SAF, are vulnerable to geometric and operating parameters. On the net effect of in-situ asphaltene removal, the alternating-injection processes would only yield higher oil recovery than the steam-alone process if there is significant in-situ deasphalting such that the oil-viscosity reduction effect overrides the permeability impairment effect. Otherwise, the miscibility of these gases in the oil-phase is not sufficiently high to take advantage of the reduction of crude viscosity by dilution. Finally, within the range of parameters evaluated, the three processes are technically and commercially feasible for the Nigerian deposit investigated. However, in terms of economics and robustness against commercial risks, the order is SAC > steam-alone > SAF. The reservoir model, oil price and costs are found to be the main determinants of project risks. Given the limitations of this research and the uncertainties in the input data used for analyses, we complete the work by outlining the scope for further studies.

This thesis reports investigations into the flotation separation of pyrite from arsenopyrite using xanthates and the new collectors Fl, S7 and F2B developed by Dow Chemical Company. The Dow surfactants were classified as chelating collectors by the manufacturer. For this purpose Hallimond tube, bench and column flotation tests were run in order to assess the flotation behaviour of these minerals. In addition to these experiments other tests performed were adsorption characteristics of Fl on pyrite and arsenopyrite, the effect of Fl on electrophoretic mobility and surface tension of Fl solutions. The effect of pH on the recovery of pyrite and arsenopyrite was studied for all collectors with the Hallimond tube. pH values ranging from 2 to 12 were tested at constant collector concentrations. At the optimum pH values further experiments were carried out under varying collector concentrations. The results obtained for five different collectors showed that selectivity was rather poor at all conditions except for Fl. A successful separation was seen to be possible employing Fl and by floating pyrite. This was proven by using a mixture of equal amounts of pyrite and arsenopyrite and selectively floating pyrite. Bench and column flotation tests were carried out in order to confirm the results obtained with the Hallimond tube. A Denver flotation cell was used and a column cell was constructed. In these tests Fl was used as a collector and in some cases amyl xanthate was blended. The effect of collector concentration was monitored and other parameters were adjusted for an optimum separation. Observations showed that a selective flotation of pyrite from arsenopyrite could be possible. Adsorption characteristics of Fl on pyrite and arsenopyrite were studied as a function of collector concentration, pH and time. Though there was not a direct correlation between the amount of Fl adsorbed and the success of flotation, in general pyrite was found to adsorb more Fl than arsenopyrite. Electrophoretic mobility of pyrite and arsenopyrite against pH and collector concentration was measured. It was observed that Fl did not have a substantial effect on the electrophoretic mobility of these minerals. The measured pK value of Fl indicated that most of the collector in solution was in molecular form above pH 9.2.

No description available.

The purpose of hydraulic fracturing is to increase the contact area of the wellbore in the reservoir to maximise production rates. For modelling purposes, the induced fracture is assumed to be of infinite or finite conductivity. The modelled fracture tends to show either features of infinite conductivity with half slope or finite conductivity with quarter slope at early time. These flow behaviours are clear indications of a stimulated well. However, observations in some post-frac well tests report a single unit slope in early time, which indicates non-fractured well response. The objective of this study is to investigate the unusual flow behaviour associated with the testing of fractured wells following a proppant frac job and address reasons for this behaviour assuming the frac job has targeted the reservoir interval of interest. This infrequent behaviour is referred to briefly in a limited number of publications but with no clear explanation. Study suggests that the controlling factors are fracture length, fracture conductivity, non-Darcy flow in the case of gas wells and the damage caused by the fracture operation including choked fracture effect and less importantly fracture face skin. This study utilizes 3-D numerical black oil and compositional simulation in single and multi-layered reservoirs containing different fluid types. A range of factors are examined that may impact the introduced fracture flow behaviour based on actual fractured well flow features found in the literature. The main fracture and reservoir parameters investigated include: fracture half-length (xf), fracture conductivity (kfwf), fracture damage including fracture choke (Sfc) and fracture face skin (Sff), non-Darcy effect, formation permeability and many others. The study also examines fractured well behaviour in naturally fractured reservoirs and gas-condensate (lean and rich) reservoirs to investigate liquid drop out effect on the induced fracture flow behaviour. It is concluded that the investigated fracture behaviour is likely to be associated with damaged fractures of short lengths and low fracture conductivity values, which often result from poorly executed frac job on the well. Knowledge obtained from the study is applied to the analysis of well tests from actual fractured wells. Understanding the flow behaviour of fractured wells is crucial to operators and service companies in evaluating the effectiveness of stimulation work performed on the well.

Impulse-based methods efficiently and accurately model high-frequency collisions of complex shapes based on the enforcement of non-penetrating constraints. It does not rely on penalty parameters nor requires the computation of penetration between bodies. This work presents a novel necessary condition for energy conservation in impulse-based methods. In previous versions of the impulse methods, such as sequential and simultaneous impulse methods, the relative velocity at the contact points after collision is directly derived from the relative velocity before collision, in a purely simultaneous or sequential manner. This work presents a novel energy tracking method (ETM), in which the relative velocities are iteratively but gradually adjusted, simultaneously modelling their interaction at each iteration. ETM ensures the energy conservation while capturing the propagation of forces during collision. The ETM is applied to model the dynamics of fragment collision in the context of fragmentation. Two approaches of fragmentation are proposed: a finite-discrete element approach, and a low cost, fragmentation pattern-based approach. The first approach models the growth of fractures using the finite element method (FEM) and advanced re-meshing technology. This finite-discrete element approach suffers from the drawback of massive computational cost. The low-cost, fragmentation pattern-based approach separate colliding bodies directly. The fragmentation pattern is generated using Weibull distribution equations, the patterns and size distributions computed using full finite/discrete element simulations and experimental results. This work investigates the influence of fragmentation on the frequency of hang-up events and on the gravity flow of rock fragments within a block caving system. Numerical results indicate that models that do not consider fragmentation tend to overestimate the frequency of hang-up accidents.

Heap leaching is an important hydrometallurgical minerals processing technique, which involves the irrigation of leaching solution through a packed bed of ore particles. It is a complex process governed by the precise chemistry of the ores and leaching solution, the biological action of bacteria that live on the rocks, and the multiphase fluid flow through and around the rocks. In order to improve the predication of recovery and the performance of the operations, a better understanding of all aspects is vital. Several studies have suggested the importance of the fluid flow and mass transport within the heaps. The smoothed particle hydrodynamics (SPH) method has a proven track record of simulating saturated flows at very small scales for porous media and is a natural tool for gaining an insight into the flows within packed beds. In this thesis we present developments to the handling of wall boundaries in SPH with an aim towards handling arbitrary geometries and improving accuracy. As well as considering the inclusion of surface tension and contact angles for simulating unsaturated flows. A volume factor correction for curved boundaries is developed that demonstrably improves accuracy of simulations, as well as detailing a 'boundary integral method' for including wall boundaries. The developed SPH simulation code is used to run a variety of unsteady unsaturated simulations through 2D packed beds, including the effects of molecular diffusion. These simulations demonstrate the important roles that saturation and capillarity play in packed beds, with unsaturated flows having enhanced transport coefficients as compared to higher saturations. These simulations also serve to highlight the fact that even at relatively long time scales, from the perspective of the flow, that the transport is not in the simple Fickian regime, which is often used when modelling heaps and other packed beds at a larger scale.

The full understanding of the initial state of petroleum reservoirs and the fluxes that lead to compositional variations have become of huge interest to the petroleum industry. The compositional variation of reservoir fluid has great commercial impact on reservoir management and field development as it affects the value of the hydrocarbon in place, what recovery mechanisms applied and the treatment process of the extracted fluid if necessary. Lateral and vertical variation in hydrocarbon density and composition between wells are observed in many oil reservoirs under appraisal. These gradations may be due to changes in reservoir filling over geological time, in which case the variations are not in an equilibrium state, or alternatively due to an equilibrium between chemical, thermal and gravity potentials. The mixing of non-equilibrium compositional distributions is affected by Darcy flows (if there is a resulting pressure gradient), gravitational overturning (if there is a density difference) and molecular diffusion. The diffusion flux may also be affected by gravitational and thermal effects. Previous work has focused primarily on convective mixing and simple models of mixing via molecular diffusion. This work focuses on the rate of mixing via molecular diffusion, including the effects of pressure and thermal diffusion, which are modelled using the thermodynamics of irreversible processes for a single phase system. The interaction of diffusional mixing and gravitational overturning is also examined. The timescales to attain steady state are analyzed as well as the resulting compositional profiles. The developed model has been validated using simple transient analytical solution proposed by Carslaw and Jaeger (1959) for the molecular diffusion flux and Gardner et al. (1962) for the natural convection process. The diffusive fluxes in our model are also validated by steady state analytical solutions for species segregating in a thermo-gravitational column. The developed model was used to analyze the experimental results obtained for two ternary mixtures of methane, n-pentane and 1-methylnapthalene; and methane, n-pentane and undecane by Ratulowski et al. (2003). Although 1-methylnapthalene and undecane have similar molar masses, the system containing 1-methylnapthalene resulted in a bigger grading (difference in mole fraction at the top and bottom of the system) than the latter. This analysis demonstrates the impact of real mixture modelling (as opposed to the case when an ideal fluid is assumed) on the segregation-mixing process. Finally, we show how the knowledge of the timescales for observed compositional variations to reach equilibrium can be used to estimate the time since a reservoir filled. The Madison formation in the LaBarge field in Wyoming, U.S.A was studied. This is an unusual gas reservoir, as non-hydrocarbons make up about 80% of the total gas composition, with methane constituting the remainder. The methane composition varies significantly, 22% at the crest of the formation to 5% near the GWC. There are several hypotheses in the literature behind the unusual gas composition and distribution in this formation (De Bruin, 2001; Stilwell, 1989; Huang et al., 2007). We use the fluid mixing model to test the various hypotheses. The results reveal that the geothermal gradient in this field is not sufficient to make the thermal diffusion and thermal convection process in this reservoir override the effect of the molecular diffusion. We conclude that the reservoir is not yet in compositional equilibrium as molecular diffusion will completely homogenize the composition variation in this field. We propose that the currently observe composition profile is as result of the formation being enriched with CO2 at approximately 3 million years ago. This timescale is contemporaneous with the volcanic activity proposed by De Bruin (2001) and Stilwell (1989).

Froth flotation is a widely used and versatile mineral processing method for concentrating metal ores. A finely ground ore feed is processed through a flotation circuit consisting of interlinked cells where mineral particles are separated from waste material by using their differences in surface properties. The layout of the circuit has been found to greatly affect the overall flotation performance. While industrial flotation circuit design has in the past relied on experience, due to the complex nature of the process, only small scale circuit optimisations using simple flotation models were reported in literature. This work proposes a new system capable of automatic generation of optimal circuit designs for any given feed. The system combines a circuit simulator containing detailed froth-phase flotation models with a robust genetic algorithm to search through possible layouts and to produce the global optimal result. An empirical model to predict the pulp phase flotation rate constant was developed, which was used together with physics-based models describing the froth recovery and entrainment factor to simulate the flotation process. Three feed models with different complexities were also developed. Comparable flotation performance was observed between a modelled 10-cell rougher circuit and the experimental results from Northparkes copper concentration plant. Through the genetic algorithm, optimal layouts were obtained for circuits consisted of 3 to 10 cells. Layouts containing only rougher cells were able to recover a maximum amount of mineral and were found optimal for smaller circuits, whilst inclusion of cleaner cells in the optimal layouts was found more beneficial for larger circuits. This circuit modelling and optimisation system was used to study the sensitivity of flotation performance and optimal layouts to variations in feed particle size and grade. The results showed strong correlations between the variables and froth phase behaviours which determined the concentrate recovery and grade, the optimal layouts were, however, robust and relatively resilient to the changes in feed conditions.