Fire Protection of Industrial Process Equipment

Abstract

Active and passive measures are two main branches in Fire Protection Engineering (FPE). The present thesis, entitled “Fire Protection of Industrial Process Equipment”, studies parts of these topics: 1. A study of the cooling efficiency of water upon impingement onto hot metal surfaces. 2. A study of industrial grade thermal insulation as a means of Passive Fire Protection (PFP). The first part of the thesis studies the cooling efficiency of water droplets impinging onto heated metal substrates. A method for studying this was developed, and measurements were performed in the temperature range from 85 °C to 400 °C, i.e. covering the boiling regimes experienced when applying water to heated objects in fires. Stainless steel and aluminum test discs (with 50 mm diameter, 10 mm thickness, and a surface roughness of Ra 0.4 or Ra 3.0) were suspended horizontally by four thermocouples, simultaneously used to record the disc temperatures. The discs were heated by a laboratory burner prior to the experiments and left to cool with and without applying 2.5, 3.2 and 3.7 mm diameter water droplets to the discs, while the disc temperatures were recorded. The droplets were generated by the acceleration of gravity from hypodermic injection needles and hit the disc center at an impingement speed of 1.5, 2.2, 3.1 and 4.4 m/s, depending on the fall heights. The water application rate was 0.022 g/s, and the discs were aligned at 0° (horizontal), 30° and 60° inclination. Based on the recorded rate of the temperature change, as well as disc mass and disc specific heat, the absolute droplet cooling effect and the relative cooling efficiency relative to complete droplet evaporation at 100 °C were obtained. Distilled water droplets were tested on both aluminum and stainless steel. Droplets of acetone solution (300 ppm and 700 ppm) and NaCl (35 g/kg) solution, emulating seawater, were tested on aluminum discs, to evaluate the influence of an active surfactant on cooling efficiency. Typically, the water-cooling efficiency was above 60% at the temperatures of boiling crisis and below 10% at temperatures above the Leidenfrost temperature. There were significant differences in the cooling efficiency as a function of temperature for the two metals investigated. There was, however, no statistically significant difference with respect to whether the surface roughness was Ra 0.4 or Ra 3.0. The droplets of higher impact speed resulted in lower cooling efficiency, especially at disc temperatures above the Leidenfrost temperature, likely due to more vigorous droplets bouncing at higher impact speeds. Larger inclination did, as expected, result in lower cooling efficiency. At temperatures associated with nucleate boiling, the water droplets with NaCl conspicuously displayed higher cooling efficiency at about 110 °C. This may be explained by the formation of small salt deposits at the disc surface, thus improving the cooling efficiency. At temperatures between 120 °C and the Leidenfrost temperature, acetone and NaCl additives did not significantly alter the cooling efficiency. Above the Leidenfrost temperature, a minor increase in cooling efficiency was observed for the acetone solutions. Overall, the additives only marginally changed the water droplet cooling efficiency. Heat fluxes in the range 250–350 kW/m2 may be expected in industrial hydrocarbon fires. According to the NORSOK S-001 standard, a firewater flux of 10 L/m2min is mandatory for protecting pressurized equipment containing hydrocarbons. At 100% efficiency, heating and boiling this water flux requires about 430 kW/m2. At temperatures associated with boiling crisis, the suggested fire water flux would be just sufficient to mitigate the expected heat fluxes. If the metal has already been heated close to, or above, the Leidenfrost temperature, this application flux is much too low. At 10% cooling efficiency, it would only be able to withdraw 43 kW/m2 from the fire-exposed surface. The simple and straightforward technique, based on the differences in cooling rate of metal discs, with and without droplet application, proved to be well suited for assessing the cooling efficiency of water droplets from 80 °C to 400 °C. The test rig also worked well for demonstrating droplet boiling regimes and water droplet cooling efficiency to fire safety engineering students and gave them valuable insight into the limited performance of water droplets cooling when applied to hot metal surfaces. The very low water droplet cooling efficiency for temperatures close to or above the Leidenfrost point underlines the importance of early detection of fire and early activation of fire water in industrial fires to prevent escalation. The fact that fire water provides increased safety for some temperature areas, but not for all, may lead to a more nuanced appreciation of this safety measure in the total risk analysis. The results also invite a discussion of other means to prevent escalation, for example lay-out based on inherent safety principles and use of passive fire protection (PFP). The second part of the thesis focuses on current industrial challenges involving insulation of pressurized equipment containing hydrocarbons. Historically, a 50 mm layer of thermal insulation, covered by an additional layer of 50 mm PFP has been applied. Experience shows that humidity from the air has wetted the thermal insulation, at areas where the temperature is below dew point, resulting in corrosion attacks. Corrosion-related incidents are among the costliest problems facing the oil and gas industry today, especially in aging facilities. According to the new standards for thermal insulation of process equipment, 25 mm spacing should be allowed between the metal object and thermal insulation, to prevent/reduce Corrosion Under Insulation (CUI). However, the new requirements will increase the total diameter of the equipment by more than 50 mm, which may not be available without major modifications. Improved knowledge about the contribution of thermal insulation as a means of PFP can be part of a solution. In order to test thermal insulation in a configuration compliant with the new standards for insulating process equipment, a prototype/mockup was built. Thereafter, a concept for small-scale testing of mockups, resembling a part of a typical hydrocarbon distillation column, with thermal insulation in accordance with the modern requirements has been developed. The second part of the thesis demonstrates a conceptual methodology for small-scale fire testing of mockups, resembling a section of a distillation column. The concept was first tested on 16 mm thick steel walls, and the mockups were exposed to a small-scale propane flame. In order to give heat flux levels in the range 250–350 kW/m2, the flame zone was optimized by controlling the air access, as well as limiting heat losses from the combustion zone. Based on the innovative and successful test concept, the performance of thermal insulation in conjunction with 16 mm, 12 mm, 6 mm and 3 mm thick steel walls was tested to check the influence of the significantly less heat sink for the thinner walls. Regardless of the tested steel plate thicknesses, about 10 minutes passed before a nearly linear steel temperature dependency versus time was observed for the exposed steel wall. Thereafter, the thinnest plates systematically showed a faster temperature increase than the thicker plates, confirming the wall heat sink effect. During these fire tests, shrinkage of the industrial thermal insulation was observed. For the most severe tests, significant destruction of the thermal insulation was evident, and there was a need for further in-depth studies of the thermal insulation behavior when exposed to high temperatures. To study thermal insulation behavior when heated, 50 mm thermal insulation cubes were heat treated (30 min holding time) at temperatures up to 1100 °C, i.e. limited by the available muffle furnace. No clear sign of melting was observed, but sintering resulted in 25% shrinkage, i.e. thickness reduction, at 1100 °C. To study this further, thermogravimetric analysis (TGA) to 1300 °C was undertaken. The TGA revealed mass loss peaks due to anti-dusting material at 250 °C and Bakelite binder loss at 460 °C. No significant mass loss occurred above 1000 °C. Differential scanning calorimetry (DSC) to 1300 °C was also undertaken to try to shed more light on the possible degradation processes involved. The DSC analysis revealed endothermic processes related to the anti-dusting material and Bakelite mass losses at the same temperatures as for the TGA. It did, however, also reveal a conspicuous endothermic peak at 1220 °C. This peak is most likely due to melting. The endothermic processes involved when heating the thermal insulation may to a large part explain the 10 min delay in steel plate temperature increase during fire testing. Overall, the tested thermal insulation also performed surprisingly well for protecting the thin steel plates through the 30 minute test period. The results show that this test concept has great potential for low-cost fire testing of other configurations, and it may serve as a setup for product development. Further research is therefore recommended to exploit these possibilities. It may also be worthwhile to study the thermal insulation breakdown mechanisms and heat transfer properties below and at breakdown temperatures. This could possibly allow the utilization of thermal insulation as a means of passive fire protection (PFP) in areas where significant cost reduction when refurbishing old process plants and oil platforms could be achieved.