Entity

Time filter

Source Type

London, United Kingdom

Gant S.E.,UK Health and Safety Laboratory | Kelsey A.,UK Health and Safety Laboratory | McNally K.,UK Health and Safety Laboratory | Witlox H.W.M.,DNV Software | Bilio M.,UK Health and Safety Executive
Journal of Loss Prevention in the Process Industries | Year: 2013

A methodology is presented for global sensitivity analysis of consequence models used in process safety applications. It involves running a consequence model around a hundred times and using the results to construct a statistical emulator, which is essentially a sophisticated curve fit to the data. The emulator is then used to undertake the sensitivity analysis and identify which input parameters (e.g. operating temperature and pressure, wind speed) have a significant effect on the chosen output (e.g. vapour cloud size). Performing the sensitivity analysis using the emulator rather than the consequence model itself leads to significant savings in computing time.To demonstrate the methodology, a global sensitivity analysis is performed on the Phast consequence model for discharge and dispersion. The scenarios studied consist of above-ground, horizontal, steady-state discharges of dense-phase carbon dioxide (CO2), with orifices ranging in diameter from 1/2 to 2 inch and the liquid CO2 stagnation conditions maintained at between 100 and 150 bar. These scenarios are relevant in scale to leaks from large diameter above-ground pipes or vessels.Seven model input parameters are varied: the vessel temperature and pressure, orifice size, wind speed, humidity, ground surface roughness and height of the release. The input parameters that have a dominant effect on the dispersion distance of the CO2 cloud are identified, both in terms of their direct effect on the dispersion distance and their indirect effect, through interactions with other varying input parameters.The analysis, including the Phast simulations, runs on a standard office laptop computer in less than 30 min. Tests are performed to confirm that a hundred Phast runs are sufficient to produce an emulator with an acceptable degree of accuracy. Increasing the number of Phast runs is shown to have no effect on the conclusions of the sensitivity analysis.The study demonstrates that Bayesian analysis of model sensitivity can be conducted rapidly and easily on consequence models such as Phast. There is the potential for this to become a routine part of consequence modelling. © 2013. Source


Lind S.J.,University of Manchester | Stansby P.K.,University of Manchester | Rogers B.D.,University of Manchester | Lloyd P.M.,DNV Software
Applied Ocean Research | Year: 2015

The high-speed impact between a body and water is an important practical problem, whether due to wave impact on a structural deck or wall, or due to a moving body such as a ship or aircraft hitting water. The very high pressures exerted are difficult to predict and the role of air may be significant. In this paper, numerical simulations are undertaken to investigate the impact of a rigid horizontal plate onto a wave crest and, in the limit, onto a flat water surface. A two-phase incompressible-compressible smoothed particle hydrodynamics (SPH) method for water and air, respectively, is applied where the water phase imposes kinematics on the air phase at the air-water interface and the air phase imposes pressures on the water at the interface. Results are compared with experimental measurements undertaken using a drop rig positioned over a wave flume so that a horizontal plate impacts the water surface in free flight. Numerical predictions of impact pressure are quite accurate; air is shown to have a significant cushioning effect for impact on to flat water and this reduces for waves as the ratio of wave height to wavelength increases. © 2014 Elsevier Ltd. Source


Kay P.J.,University of Cardiff | Bowen P.J.,University of Cardiff | Witlox H.W.M.,DNV Software
Journal of Loss Prevention in the Process Industries | Year: 2010

This paper describes the results of the first stage of Phase III of a Joint Industry Project (JIP) on liquid jets and two-phase droplet dispersion. This stage included scaled experiments for water, gasoline, and cyclohexane for a range of superheats and nozzles with different aspect ratios. Additional experiments for butane and propane were conducted as a validation exercise and are discussed in the companion paper. Moreover this paper provides recommendations for atomisation correlations in the regimes of mechanical break-up, transition to flashing, and fully flashing. The objectives of this scaled experimental programme are to: (i) provide confidence in the previously proposed modelling methodology (Phase II) across a broad range of initial conditions (ii) update the models' correlations to generalise further its applicability (iii) recommend further model improvements. Development of new correlations for Sauter Mean Droplet diameter (SMD) and droplet size distribution is based on a best fit of the current scaled experimental data. The new data endorses the previous tri-functional Phase II approach including regimes for mechanical break-up, transition to flashing, and fully flashing, which is hence updated in the new Phase III SMD model. Considerable effort is devoted to capturing the full droplet size range under low-superheat conditions. Also, new enhancements in PDA technology were adopted to enable better quality data under high-superheat conditions. The priority recommendation for further model improvement is better characterisation of the poor quality releases under low-superheat conditions, where this work indicates that extremely broad droplet size distributions are likely. A companion paper (Part I) presents a more general overview of the dispersion problem, implementation of the correlations and subsequent performance against both the current scaled experiments and additional large-scale butane experiments. © 2010 Elsevier Ltd. Source


Witlox H.W.M.,DNV Software | Harper M.,DNV Software
Journal of Loss Prevention in the Process Industries | Year: 2013

Many accidents involve two-phase releases of hazardous chemicals into the atmosphere. This paper describes the results of the fourth phase of a Joint Industry Project (JIP) on liquid jets and two-phase droplet dispersion. The objective of Phase IV of the JIP was to generate experimental rainout data for non-flashing experiments, and to develop recommendations for the best methodology to predict rainout [total rainout mass and its spatial distribution ('distributed' rainout)]. Phase IV of the JIP first included rainout experiments by the UK Health and Safety Laboratory (HSL) for sub-cooled releases of water and xylene with a range of orifice sizes and stagnation pressures. See the companion paper II for further details. Secondly model validation was carried out by DNV Software for these experiments using different correlations for the initial droplet size (Sauter Mean Diameter, SMD), i.e. the CCPS SMD correlation and the Phase III JIP SMD correlation. The validation includes flow rates, droplet size, distributed rainout and cloud temperature drop. Subsequently validation was considered for a wider range of experiments from the literature (sub-cooled and superheated releases) for both SMD and total rainout. Adopted rainout methods comprised both methods including explicit modelling of the droplets (using an extended version of Phast dispersion model UDM), as well as more simple methods based on rainout correlations without droplet modelling. Recommendations are made for the most accurate droplet size and rainout modelling. A modified CCPS UDM droplet size correlation has been shown to agree best against experimental rainout data. © 2012 Elsevier Ltd. Source


Bettis R.J.,HSL Fire and Process Safety | Jagger S.F.,HSL Fire and Process Safety | Witlox H.W.M.,DNV Software | Harper M.,DNV Software
Journal of Loss Prevention in the Process Industries | Year: 2013

This paper describes the results of the first experimental stage of Phase IV of a Joint Industry Project (JIP) on liquid jets and two-phase droplet dispersion. The objective of this stage of the JIP was to generate experimental rainout data for non-flashing water and xylene experiments. See the overview companion paper I for a wider overview of the problem, model implementation and associated model validation.A range of orifice sizes (2.5 and 5 mm) and stagnation pressures (4-16 barg) were applied. Measurements included flow rate, initial droplet size, plume concentrations/temperatures for a range of downstream locations, and distributed rainout. Instead of the Phase Doppler Anemometry method used for droplet size measurements earlier in the JIP, a photographic technique was applied in an attempt to include measurement of the larger (non-spherical) droplets. This enabled a more accurate evaluation of the initial droplet size distribution and a much clearer understanding of the droplet morphology. The results showed that the droplet behaviour in the jet is more complex than had been anticipated with the mass distribution dominated by a very small number of large non-spherical droplets. Consequently a large number of spray images were required to evaluate an accurate size distribution. Distributed rainout was measured by weighing the amount of rainout in trays positioned along the jet direction. The rainout results showed a good degree of repeatability and internal consistency. They indicated that an increasing proportion of the released material did not rainout for increasing pressure. Rainout distance also increased with increasing pressure. Evaporation of the liquid was confirmed by temperature measurements, which showed the effect of evaporative cooling. Xylene concentration measurements (up to 1%) were carried out using a direct reading photoionization detector calibrated for xylene (measuring vapour only). For a limited dataset, the accuracy of these measurements was estimated by means of comparison against an alternative more time-consuming concentration method (xylene absorption onto a charcoal filter; measuring both vapour and liquid). The concentration measurements displayed several consistent qualitative features. For example, at a given downstream distance, the peak concentration increases with increasing pressure and nozzle diameter and the vertical height at which the peak is achieved increases. The cross-stream profiles displayed a consistent tendency to increased concentration at the edge of the jet, and the reason for this has not been established. Finally recommendations are provided for potential future work. © 2012 Elsevier Ltd. Source

Discover hidden collaborations