• / 80
  • 下载费用:5 下载币  

TNO多能爆炸模型指南 RIGOS

关 键 词:
定量计算 dnv 事故 概率
资源描述:

HSE Health  · the type of fuel; · the geometric scale. To parameterise an obstacle configuration, three parameters were introduced: · The Volume Blockage Ratio (VBR) which is the ratio of the summed volume of the obstacles in an obstructed region and the volume of that region, assuming the obstacles consist of cylinders; · The distance a flame can propagate within an obstructed region (L f ); · The average obstacle diameter (D). 3 TNO report COMPANY CONFIDENTIAL 8 L PML 2002-C50 The influence of the scale and the fuel is taken into account by a theory, developed by Taylor and Hirst (1988), Catlin (1991) and Catlin and Johnson (1992), based on Karlovitz number similarity. Because the origin of scale effects in gas explosions is predominantly in the scale of the turbulence, the average obstacle diameter D was chosen as the scale parameter while the laminar burning velocity S l was adopted as the parameter characteristic for the reactivity of the flammable mixture. 1 10 Hj 0.001 0.01 0.1 100 Overpres s u re ( B ar) Harrison ertager MERGE -2.50 -2.00 -1.50 -1.00 -0.50 0.00 0.50 1.00 b 2.7 0.7 LOG ((VBR*L f /D) *Sl *D ) Figure 1: Observed overpressures and correlation line for MERGE experiments. The correlation, consisting of a best fit of the experimental data in Figure 1, is expressed as: 75 , 2 .L VBR f öæ 7.2 7 .0 P ç ç è 84.0 .S D÷ ÷ D = l . o D ø The correlation makes it possible to make an estimate of an explosion overpressure in realistic problems such as, for instance, a chemical or refinery plant. The follow-up project GAMES (Mercx, 1998) was meant to investigate the practi- cal difficulties encountered when the correlation was applied to a realistic plant. The exercises performed in GAMES resulted in practical guidance for the determi- nation of the volume blockage ratio VBR, the flame path length L f and the average obstacle diameter D. 2.2 The Critical Separation Distance The Multi-Energy method recognises that in gas deflagration, turbulence genera- tive boundary conditions are the predominant factor in the development of over- pressure and blast. The mechanism of a gas deflagration implicates that as soon as the appropriate turbulence generative boundary conditions are lacking, the burning 4 TNO report COMPANY CONFIDENTIAL 9PML 2002-C50 speed and the pressure build-up in the process of flame propagation drop. The implicit assumption is that the flammable mixture in the cloud is too inhomogene- ously mixed to maintain a detonation. The direct consequence of the Multi-Energy concept is that an extended vapour cloud containing several obstructed areas, separated by open spaces of sufficient extent, will produce the same number of separate blast waves on ignition. In the modelling of blast effects, therefore, the individual obstructed areas should be separately considered. The problem has been visualised in Figure 2. A large flammable vapour cloud has covered two densely obstructed areas of explosive potential: a chemical plant and closely parked boxcars at a railway shunting yard. The space in between the two regions is open and unobstructed. If the distance between the two is sufficient, the ensuing gas explosion on ignition will develop two separate blasts. Figure 2: Two obstructed regions in a large cloud. One big or two smaller explosions? The Yellow Book (CPR-14E, 1997) intuitively defines a congested area as an area in which obstacles are positioned within a 10 obstacle diameters distance from one another with an upper limit of 25 m. This statement would suggest a Critical Sepa- ration Distance equal to 10 obstacle diameters with an upper bound of 25 m. In the final report of the GAMES-project, a preliminary safe and conservative guideline for the quantification of the term “open spaces of sufficient extent” has been developed on the basis of global theoretical considerations. This preliminary guideline runs as follows: · The Critical Separation Distance around a potential blast source area is equal to half its linear dimension in each direction. · If the distance between potential sources is larger, the sources should be mod- elled as separate blasts. · If not, they should be modelled as one single blast of summed energy content. 5 TNO report COMPANY CONFIDENTIAL 10PML 2002-C50 This preliminary guidance was based on the observation that the combustion proc- ess in a gas explosion is driven by turbulence. As the turbulence is generated in the expanding medium in interaction with the obstacles, the influence of the turbulent combustion cannot extend beyond the fuel-air mixture initially present within the obstructed area. As the expansion factor of stoichiometric fuel-air mixtures is approximately equal to 8, the turbulent combustion in a threedimensionally ex- panding gas explosion will be anyway limited to a volume of a linear dimension equal to twice that of the obstructed region. It is to be expected that the critical separation distance can be expressed as a certain portion of the size of the obstructed donor volume, indeed, but can probably be taken smaller than the preliminary estimate. The obstacles in an obstructed area determine the scale of the turbulence and thereby the turbulence decay outside the obstructed region. Because turbulence is the predominant combustion-driving phenomenon in gas explosions, it may be expected that the critical separation distance is also dependent on the diameter of the obstacles. 6 TNO report COMPANY CONFIDENTIAL 11PML 2002-C50 3 Research program 3.1 Objectives The primary objective of the project is to develop practical guidance with regard to the Critical Separation Distance, a basic element in the application of the Multi- Energy method. This overall objective can be split into several related sub- objectives, which can be designated as follows: · To generate a database on donor-acceptor gas explosions. · To generate understanding of the process of flame propagation from donor to acceptor in relation to the blast produced. · To measure the Critical Separation Distance dependent on the donor size, the donor obstacle density and fuel reactivity. · To investigate whether the Critical Separation Distance is influenced by the diameter of the obstacles in the donor obstacle configuration. · To investigate whether the Critical Separation Distance is influenced by a connection of small cross-sectional dimension between two obstacle configura- tions, representing a pipe rack between process units of a chemical plant. · To develop practical guidance on the Critical Separation Distance in the appli- cation of the Multi-Energy method. The availability of the substantial body of experimental data generated in this project offered the opportunity to extend the program with an additional objective, namely: · To evaluate the performance of simple vapour cloud explosion blast modelling methods on explosions substantially differing from those, these methods were derived from. 3.2 Approach To determine the Critical Separation Distance, two configurations of obstacles were placed at a certain distance (Separation Distance) from one another. The two configurations including the open space in between were enclosed in plastic sheet to contain a flammable gas mixture. One obstacle configuration, in whose centre the flammable cloud was ignited, was referred to as the ‘donor’ and the other as the ‘acceptor’. In a series of tests the Separation Distance was varied and the blast was measured at various locations around. The maximum separation distance at which the blast waves from donor and acceptor were found to coincide, was designated as the Critical Separation Distance. The size and obstacle density of the acceptor were kept constant all over the pro- gram as they were considered not to influence the Critical Separation Distance. The 7 TNO report COMPANY CONFIDENTIAL 12PML 2002-C50 composition and dimensions of the donor, on the other hand, were expected to have a substantial influence. Therefore, the following parameters were varied: · The dimensions of the Donor (DD); · The Volume Blockage Ratio of the donor (VBR); · The Fuel (F); · The Separation Distance between donor and acceptor (SD). In some additional tests, it was investigated if the obstacle diameter in the donor has some influence on the Critical Separation Distance. Besides, a very limited number of experiments have been performed in which the donor and acceptor were connected by an obstacle configuration of small cross-sectional dimensions, repre - senting a pipe rack between separate process units of a chemical plant. Both, the size of the connecting obstacle configuration as well as the fuel were varied. 3.3 Test location The experiments have been performed at the so-called FAST facility of TNO Prins Maurits Laboratory. The facility consists of a concrete pad in which various cable trays are present for the protection of cables and other measuring equipment. The concrete pad is situated in an open flat terrain, large enough to prevent reflections from influencing the measurements. 3.4 The obstacle configurations The composition of the obstacle configurations are identical to those used in the MERGE and EMERGE projects. They consist of a number of tubes of circular cross-section. The tubes are orientated in a fully regular way in three perpendicular directions. Figure 3 shows a part of a typical MERGE obstacle configuration. This particular part has 5 horizontal layers of cylindrical obstacles. Each layer consists of 5 by 5 obstacles orientated in two perpendicular directions. There are 25 vertical obstacles; each of these connects the knots in the horizontal layers at corresponding horizontal co-ordinates. This particular configuration is denoted as a 5 ´ 5 ´ 5 obstacle configuration. 8 TNO report COMPANY CONFIDENTIAL 13PML 2002-C50 Figure 3: Typical MERGE obstacle configuration as used in the RIGOS-project. The composition of an obstacle configuration is characterised by: · the Diameter of the cylinders (D); · the axial spacing (Pitch) between adjacent tubes (P); · the Number of cylinders in a row (N). The length L and the width W of the configurations were taken equal, while the height H was taken as 0.5L. The ignition location was in the centre of the donor at ground level. A tube diameter of D=19.1 mm in arrays of pitches of P = 4.65D and P = 7D resulted in obstacle configurations of volume blockage ratios of VBR = 10.1% and VBR = 4.6% respectively. The obstacle configurations were indicated as type A and type B, respectively. The configuration of the acceptor is characterised by N = 16, P = 4.65D, giving a VBR = 10.1% and was kept constant all over the program. 3.5 The test set-up 3.5.1 Layout Roughly speaking, all experiments were of a similar set-up as represented in Figures 4a and 4b. 9 TNO report COMPANY CONFIDENTIAL 14PML 2002-C50 F9 P9 F1 P8P7P6P5P4P3 P2 P1B1 3m from centre of donor B2, B3 and B4 3,6 and 12 m from centre of acceptor Figure 4a: Test set-up in the AE, AM, BE and BM test series. Experiments were performed on a concrete pad in which a cable tray has been cut away. The cable tray was covered with a steel lid in which measuring instrumenta- tion can be mounted. The donor and acceptor obstacle arrays were placed on the pad centred on the cable tray (Figure 4a and 4b). Together they were enclosed in a tent of plastic sheet to contain the flammable gas mixture. The position of the acceptor was fixed all over the experimental program. The position of the donor varied with the variation of the separation distance and the donor size. Pressure was measured in 9 stations positioned at more or less regular distances along the axis within the donor-acceptor configurations (P1 – P9). The pressure gauges P4 to P9 were mounted in and flush to the cable tray lid. The pressure gauges P1 to P3 were fixed to the donor obstacle array. At nearly the same loca- tions, thermocouples were mounted to measure flame arrival times (F1-F9). In the first 4 tests AE01 – AE04, blast overpressures were recorded at 1 station at 3-m distance from the donor centre (B1), at 2 stations at 3 and 6-m distance from the acceptor centre in the donor-acceptor direction and at 1 station at 3-m distance from the acceptor in cross direction. In the rest of the AE, AM, BE and BM test series, the blast overpressures were measured at 1 station at 3-m distance from the donor centre (B1) and 3 stations at 3, 6 and 12 m distance respectively from the acceptor centre (B2-B4). To study directionality effects in the blast, in a later stage in the program the num- ber of blast overpressure gauges was extended. As indicated in Figure 4b, the number of pressure gauges near the donor was extended to 3, positioned at 3, 6 and 12 m distance from the donor centre. The number of pressure gauges near the acceptor was extended with 3, positioned at 3, 6 and 12 m distance from the ac- ceptor centre perpendicular to the direction of flame propagation (Figure 4b). B1,B2 and B3 3,6 and 12 m from centre of donor F9 P9 F1 P8P7P6P5P4P3 P2P1 B4, B5 and B6 3,6 and 12 m from centre of acceptor B7, B8 and B9 3,6 and 12 m from centre of acceptor Figure 4b: Test set-up in the AP, CP, DP and the DM test series. 10 TNO report COMPANY CONFIDENTIAL 15PML 2002-C50 3.5.2 Instrumentation The signals from the pressure gauges and thermocouples have been transmitted to the SCADAS II Signal Conditioning and Data Acquisition System. Pressures inside the donor-acceptor configuration are measured by means of piezo-resistive transducers, Druck type PDCR 200 (FS 600 kPa). The membranes are covered with 2 mm black and 1 mm white silicon grease to reduce flashlight sensitivity and drift caused by the heat of the flame. Two different kinds of blast gauges have been be used: blast pencils (Kulite trans- ducers) as well as pressure gauges mounted in skimmer plates (Druck PDCR 10/F transducers). To detect the flame position during the experiments Chromel/Alumel thermocou- ples have been used. Thermocouples record a temperature increase and thereby the passage of a flame front (interface between unburned and burnt gases). The ignition consisted of a spark capacitor circuit and a spark plug. The energy released to the spark plug has a maximum gross content of 2 Joule. A combination of electromagnetic and optic techniques determines the moment of ignition. 3.5.3 Mixture preparation and control The flammable gas is taken from a gas cylinder and piped into the array through a gas inlet. Fans at the perimeter of the obstacle array create a flow inside the array to promote adequate mixing of gas and air. The fans are switched off about one minute prior to ignition. The gas mixture is pumped to the gas analyser via suction lines. One sample point is a few cm above ground level and a second sample point is located near the top of the array. Samples are taken continuously at the two locations until one minute prior to ignition. The samples are inserted to an Infrared Analysis System. The concentrati

展开阅读全文
  石油文库所有资源均是用户自行上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作他用。
0条评论

还可以输入200字符

暂无评论,赶快抢占沙发吧。

关于本文
本文标题:TNO多能爆炸模型指南 RIGOS
链接地址:http://oilwenku.com/p-70631.html
关于我们 - 网站声明 - 网站地图 - 资源地图 - 友情链接 - 网站客服客服 - 联系我们
copyright@ 2016-2020 石油文库网站版权所有
经营许可证编号:川B2-20120048,ICP备案号:蜀ICP备11026253号-10号
收起
展开