This document provides a description of the model evaluation protocol (MEP) database for fires involving liquefied natural gas (LNG) and processing fuels at LNG facilities. The purpose of the MEP is to provide procedures regarding the assessment of a model's suitability to predict thermal exclusion zones resulting from a fire. The database includes measurements from pool fire, jet fire, and fireball experiments which are provided in a spreadsheet. Users are to enter model results into the spreadsheet which automatically generates statistical performance measures and graphical comparisons with the experimental data. The intent of this document is to provide a description of the experiments and of the procedure required to carry out the validation portion of the MEP. In addition, the statistical performance measures, measurements for comparisons, and parameter variation are provided.
This document provides a description of the model evaluation protocol (MEP) for pool fires, jet fires, and fireballs involving liquefied natural gas (LNG), refrigerant fluids, and byproducts at LNG facilities. The purpose of the MEP is to provide procedures regarding the assessment of a model's suitability to predict heat flux from fires. Three components, namely, a scientific assessment, model verification, and model validation comprise the MEP. The evaluation of a model satisfying these three components is to be documented in the form of a model evaluation report (MER). Discussion of models for the prediction of fire, detailed information on each of the three MEP components, the MEP procedure regarding new versions of previously approved models, and the format of the model evaluation report (MER) are provided.
This report provides results from a series of 2-m pool fire experiments performed in the Thermal Test Complex at Sandia National Laboratories testing heptane, Bakken crude oil, and dilbit crude oil. The effect of the presence and placement of a calorimeter, fuel supply temperature, and maintaining a constant fuel level were assessed. Measurements include burn rate, surface emissive power, flame height, heat flux to an engulfed calorimeter, heat flux to external instruments, thermocouple temperatures within the fuel and fire plume, and heat release rate. The results indicate that the presence and placement of the calorimeter has the most effect on the measured quantities for the Bakken crude oil and indicated no effect for the Dilbit crude oil. The fuel feed temperature had a slight effect for the heptane fuel, but not for the crude oils. Allowing the fuel to burn down did not have a significant effect on any of the fuels. The Bakken crude oil resulted in the highest average total heat flux to the calorimeter by a factor of about 1.5 and 1.3 higher compared to heptane and the dilbit crude oil, respectively.
In 2004, at the request of the Department of Energy, Sandia National Laboratories (Sandia) prepared a report, "Guidance on the Risk and Safety Analysis of Large Liquefied Natural Gas (LNG) Spills Over Water". That report provided a framework for assessing hazards and identifying approaches to minimize the consequences to people and property from an LNG spill over water. Because of increasing domestic U.S. supplies of natural gas and associated by products, such as liquefied propane gas (LPG), the United states Coast Guard requested that Sandia assess the general scale of possible hazards for a breach and spill of an LPG carrier. Because of the broad range of LPG carriers types -- refrigerated and pressurized, ships and barges, Sandia chose to focus this analysis on the larger LPG refrigerated systems. With cargo capacities ranging up to 100,000 m 3 , these types of ships can be expected to support potential increased LPG exports. Sandia assessed potential accidental and intentional threats, and based on LPG carrier configurations and designs, estimated potential breach sizes, spill rates and volumes, and conducted fire, vapor dispersion, and detonation hazard analyses. This report summarizes the analyses conducted, the expected range of potential hazards from an associated refrigerated LPG carrier spill over water, and risk management approaches to minimize consequences to people and property from such a spill.
A series of experiments were performed with the objective of achieving an extreme thermal environment by creating a fire whirl in an enclosure in facilities at the Thermal Test Complex (TTC) at Sandia National Laboratories. The motivation for the experiments is based on results from previous experiments performed at Sandia in an igloo representing a mock weapon's storage facility. In that test series, a fire whirl developed within the igloo resulting in extremely high heat flux levels. This environment was created with a pool fire of 4.6-m in diameter and was not under controlled, repeatable conditions. The objective of the current tests is to have the ability to create this environment in a repeatable controlled environment at a smaller scale, namely with a pool fire not above 3-m diameter effectively, thereby allowing for repeatable, cost-effective testing. In FY15, six tests were conducted in the Crosswind Test Facility (XTF), using a 1.77 m square pan. In FY16, three tests were conducted in the Fire Laboratory for Accreditation of Modeling by Experiment (FLAME) using a 3-m diameter pan. Both of these test series utilized the same enclosure. In FY17, a single test was performed in XTF using a 2.7 m square pan using a modified enclosure which included a ceiling. All tests used Jet-A as the fuel. The wind speed and gap width of the enclosure were varied for the FY15 XTF tests and the gap width and effect of insulation on the enclosure walls were varied for the FY16 FLAME tests. Fuel regression rates, heat flux, and gas velocity measurements were obtained. The results from the FY15 and FY16 test series indicate that fuel regression rates and peak heat flux levels are a factor of two higher than non-fire whirl pool fires of equivalent diameter. The results from the FY17 test using an enclosure with a ceiling met the objective of the test series by achieving temperatures of nearly 1400degC and heat flux levels of 400 kW/m 2 .
The objective of this work is to assess dispersion distances of a vapor mixture of species released from a railcar containing a tight crude oil. Tight crude oils can have higher levels of light ends as compared to conventional crude oils [1], which if released and dispersed could pose a potential hazard with regards to a flash fire, explosion, and/or asphyxiation. A historical accident involving rail transport in Viareggio, Italy illustrates how the spillage of LPG can lead to severe damage as a result of a propagating vapor cloud [2]. One of 14 railcars was punctured after derailment, releasing about 110 m3 of LPG into a densely populated area (2000 persons/km2 ). The resulting vapor cloud propagated and infiltrated nearby buildings and houses which were an average of 10 m in height. Ignition of the cloud occurred approximately 100 to 300 seconds after the start of the spill. A flash fire and explosions resulted, killing 31 people. Evidence suggests that most deaths occurred due to the asphyxiation and thermal hazards from the flash fire. Thus, the motivation for this work is to assess if significant vapors can develop from a railcar carrying a tight crude oil and if this cloud could disperse potentially to nearby populations.
At the request of GDF Suez, a Rough Order of Magnitude (ROM) cost estimate was prepared for the design, construction, testing, and data analysis for an experimental series of large-scale (Liquefied Natural Gas) LNG spills on land and water that would result in the largest pool fires and vapor dispersion events ever conducted. Due to the expected cost of this large, multi-year program, the authors utilized Sandia's structured cost estimating methodology. This methodology insures that the efforts identified can be performed for the cost proposed at a plus or minus 30 percent confidence. The scale of the LNG spill, fire, and vapor dispersion tests proposed by GDF could produce hazard distances and testing safety issues that need to be fully explored. Based on our evaluations, Sandia can utilize much of our existing fire testing infrastructure for the large fire tests and some small dispersion tests (with some modifications) in Albuquerque, but we propose to develop a new dispersion testing site at our remote test area in Nevada because of the large hazard distances. While this might impact some testing logistics, the safety aspects warrant this approach. In addition, we have included a proposal to study cryogenic liquid spills on water and subsequent vaporization in the presence of waves. Sandia is working with DOE on applications that provide infrastructure pertinent to wave production. We present an approach to conduct repeatable wave/spill interaction testing that could utilize such infrastructure.
Effective capture of atmospheric carbon is a key bottleneck preventing non bio-based, carbon-neutral production of synthetic liquid hydrocarbon fuels using CO{sub 2} as the carbon feedstock. Here we outline the boundary conditions of atmospheric carbon capture for recycle to liquid hydrocarbon fuels production and re-use options and we also identify the technical advances that must be made for such a process to become technically and commercially viable at scale. While conversion of atmospheric CO{sub 2} into a pure feedstock for hydrocarbon fuels synthesis is presently feasible at the bench-scale - albeit at high cost energetically and economically - the methods and materials needed to concentrate large amounts of CO{sub 2} at low cost and high efficiency remain technically immature. Industrial-scale capture must entail: (1) Processing of large volumes of air through an effective CO{sub 2} capture media and (2) Efficient separation of CO{sub 2} from the processed air flow into a pure stream of CO{sub 2}.
An uncertainty quantification (UQ) analysis is performed on the fuel regression rate model within SIERRA/Fuego by comparing to a series of hydrocarbon tests performed in the Thermal Test Complex. The fuels used for comparison for the fuel regression rate model include methanol, ethanol, JP8, and heptane. The recently implemented flamelet combustion model is also assessed with a limited comparison to data involving measurements of temperature and relative mole fractions within a 2-m diameter methanol pool fire. The comparison of the current fuel regression rate model to data without UQ indicates that the model over predicts the fuel regression rate by 65% for methanol, 63% for ethanol, 95% for JP8, and 15% for heptane. If a UQ analysis is performed incorporating a range of values for transmittance, reflectance, and heat flux at the surface the current model predicts fuel regression rates within 50% of measured values. An alternative model which uses specific heats at inlet and boiling temperatures respectively and does not approximate the sensible heat is also compared to data. The alternative model with UQ significantly improves the comparison to within 25% for all fuels except heptane. Even though the proposed alternative model provides better agreement to data, particularly for JP8 and ethanol (within 15%), there are still outstanding issues regarding significant uncertainties which include heat flux gauge measurement and placement, boiling at the fuel surface, large scale convective motion within the liquid, and semi-transparent behavior.
This paper applies a pragmatic approach to validation of a fire-dynamics model involving computational fluid dynamics, combustion, participating-media radiation, and heat transfer. The validation problem involves experimental and predicted steady-state temperatures of a calorimeter in a wind-driven hydrocarbon pool fire. Significant aleatory and epistemic sources of uncertainty in the experiments and simulations exist and are transformed to a common basis of interval uncertainty for aggregation and comparison purposes. The validation comparison of experimental and simulation results, and corresponding criteria and procedures for model substantiation or refutation, take place in "real space" as opposed to "transform space" where various transform measuresof discrepancy between experiment and simulation results are calculated and assessed. The versatile model validation approach handles difficulties associated with representing and aggregating aleatory and epistemic uncertainties (discrete and continuous) from multiple correlated and uncorrelated source types, including 1) experimental variability from multiple repeat experiments, 2) uncertainty of experimental inputs, 3) experimental output measurement uncertainties, 4) uncertainties that arise in data processing and inference from raw simulation and experiment outputs, 5) parameter and model-form uncertainties intrinsictothe model, and 6) numerical solution uncertainty from model discretization effects. Copyright Clearance Center, Inc.