Special Problems in Fire Protection Engineering

Paul DeCicco, Paul R DeCicco

• Publisher: Routledge
• Year: 2019
• Print ISBN: 9780895032232
• eBook ISBN: 9781351854559
• Topic: Psychology
• Subtopic: Environmental Management
• Index: Psychology

Features papers directed to fire protection in various environments other than building structures including fuel transporting vehicles, spacecraft, a sports arena, an offshore oil rig and propane fueling bus facilities.

---

CHAPTER 1
Modelling the Effects of Temperature Sensitive Pressure Relief Devices on Tankers Exposed to Engulfing Fires

A. Michael Birk

The present research involved the use of the Tank-Car Thermal Computer Model (TANKCAR) [1] to model the effects of fusible plug type temperature sensitive pressure relief valves (TPRVs) on an uninsulated tank. TANKCAR is a digital computer-based model that can simulate the thermal response of a long cylindrical tank filled partially with liquid and partially with vapor exposed to either an engulfing type fire, such as caused by a burning pool, or a torch type fire, such as that caused by a relief valve flare from a neighboring tank. The model can account for the effects of a number of thermal protection devices such as pressure relief valves, and novel internal protection devices including heat dissipating matrices. The model can simulate the effects of roll and pitch of the tank. Simulation results are presented for cases involving different numbers of fusible plug type TPRVs, and cases involving a fixed number of fusible plugs of different melting temperatures. In all cases the tanks are uninsulated with standard sized pressure relief valves and are filled with propane. The simulations involve engulfing type fires. The simulation results indicate that TPRVs could be very effective at reducing the risk of thermal ruptures of tanks. The TPRV simulation results are compared with experimental data and simulation results for an uninsulated tank filled with propane exposed to an engulfing fire.

Rail and highway tankers are equipped with thermal protection systems to minimize the effects of external heating such as may be caused by accidental fire impingement. Typical thermal protection systems include pressure activated pressure relief values (PPRV) and thermal insulation [2]. The PPRV’s maintain the tank internal pressure below some safe value, and the thermal insulation material acts as a barrier to heat addition. The thermal insulation has the dual effect of acts as a barrier to heat addition. The thermal insulation has the dual effect of maintaining wall temperatures below some critical value, and reducing the vapor generation rate within the tank [2].

The PPRV/thermal insulation combination has proven to be very effective. However, if for some reason the thermal insulation layer is damaged, and some of the tank wall is exposed directly to the effects of fire impingement, the wall temperatures in that region can increase dramatically. The resulting increase in wall temperatures can result in significant degradation of the tank wall material properties, and as a consequence the pressure carrying capability of the tank can be reduced. If the wall temperatures are allowed to go high enough, a thermal rupture of the tank will occur even if the PPRV is operating properly [2] to maintain the tank pressure below the nominal burst pressure of the tank.

One possible solution to this hypothetical situation is the temperature sensitive pressure relief valve (TPRV). This device acts to further reduce the tank internal pressure if high wall temperatures are sensed. Such a device would be sized to reduce the tank internal pressure well below that possible by the PPRV.

The TPRV can take different forms. Simple TPRV’s can be valves with temperature sensitive elements that result in the opening of an orifice when the temperature increases. In the case of fusible plug type devices, a low melting temperature plug melts and is blown clear to render an orifice for pressure relief. More complex devices can be envisaged which involve data collection from strategically placed thermocouples, and the appropriate mechanisms to open a valve if temperatures exceed some critical value.

The present research involved computer simulations of tanks equipped with TPRV’s filled with propane, exposed to engulfing type fires. The simulations were carried out using the tank-car thermal analysis program TANKCAR with modifications so that it could account for the effects of the TPRV’s.

THE MODEL

The model used in the present research is based on the computer program TANKCAR [1]. This computer model can simulate a long cylindrical tank filled partially with liquid and partially with vapor exposed to either an engulfing type fire, such as caused by a burning pool, or a torch type fire, such as that caused by a relief valve flare from a neighboring tank. The model can account for the effects of a number of thermal protection devices such as pressure relief valves, thermal insulation, radiation shielding, temperature sensing relief valves, and novel internal protection devices, including heat dissipating matrices. The model is basically a two-dimensional representation of a circular cylindrical tank (i.e., axial gradients and end effects are not accounted for). However, the model has a pseudo-3D operating mode so that pitched and rolled tanks can be analyzed.

TANKCAR is capable of predicting the tank internal pressure, mean lading temperatures, wall temperature distribution, relief valve flow rates, liquid level, tank wall stresses and tank failure all as functions of time from initiation of the fire impingement. These various outputs define the response of the tank/lading system and provide valuable information for the design of a thermal protection system.

The model has been extensively validated [3] by comparing its predictions with the results of numerous fire tests involving full [4] and fifth-scale [5, 6] rail tank-cars exposed to engulfing fires.

TANKCAR is made up of a series of submodels simulating the following processes:

1. flame to tank heat transfer,

2. heat transfer through the tank wall and associated coverings,

3. interior-surface to lading heat transfer,

4. thermodynamic process within the tank,

5. thermodynamic properties of the lading,

6. pressure relief device operating characteristics,

7. wall stresses and material property degradation, and

8. tank failure.

The flame-to-tank heat transfer submodels can account for either an engulfing pool fire or a two-dimensional torch. The pool fire model accounts for both thermal radiation and free convection heat transfer from the fire to the tank. The thermal radiation is calculated as a function of the circumferential position on the tank surface by accounting for the typical shape of large pool fires in the absence of cross wind effects. The convection calculations are based on empirical relations for convection to horizontal cylinders in a crossflow. The torch submodel is based on empirical relations for the heat transfer from a jet impinging on a flat plate. The geometry of the jet/plate system in the submodel is similar to the U.S. DOT Transportation Test Center Torch Simulator as described in Reference [7].

The heat transfer through the tank wall and associated coverings is represented using finite difference techniques. The finite difference solution accounts for pure conduction through the tank wall and insulating layers, and pure conduction, convection, and radiation when appropriate, such as in the case of certain heat conducting matrices. The finite difference solution also accounts for the heat transfer from the fire to the tank outer surface and from the tank inner surface to the lading. The interior surface heat transfer submodels account for convection and radiation in the vapor space, and convection and boiling in the liquid region. The radiation calculations are based on the assumption that the wall communicates only with the lading. The convection and boiling coefficients are based on empirical relations for inclined flat surfaces.

The thermodynamic process submodel treats the lading as three distinct regions—the vapor space, the liquid boundary, and the liquid core. It is assumed that the vapor and liquid boundary (near wall, and free surface) are in thermodynamic equilibrium and saturated; the core is assumed to be subcooled initially, but after some period of venting it is assumed to be in equilibrium with the liquid boundary and vapor space. This submodel requires the setting of two empirical constants, the liquid boundary thickness, and the energy partition factor that determines how much of the fire heat is transferred into the vapor and liquid boundary, and how much is transferred to the liquid core. These constants have been determined by calibrating the model with one set of fire test results (full-scale tank fire test [5]) and have remained unchanged for all subsequent validation runs and simulations. In the submodel it is assumed that the energy for venting comes from the vapor and liquid boundary. This energy drain is one reason that the liquid core eventually reaches equilibrium with the other regions.

The thermodynamic and transport properties of the lading are based on the Starling equation of state [8], and on available material property data, respectively. The pressure relief valve submodel accounts for both the mechanical action of a relief valve and the fluid mechanics. The valve mechanics are accounted for using a steady state model. Valve cycling dynamics are not accounted for explicitly, but rather implicitly by having a model valve that remains partially open to represent the reduced flow capacity of a real valve during cycling, and fully open when in reality the valve is fully open. The opening fraction is determined by the lading vaporization rate. The valve flow submodels can account for choked dry vapor flow and for frozen liquid flow.

The wall stresses are calculated at the point on the wall circumference which experiences the highest temperature and includes pressure induced (hoop) stress and stresses due to radial temperature gradients in the tank wall. The tank failure analysis is based on the maximum normal stress theory of failure. Degradation in the wall material strength with increases in temperature is accounted for using available data for tank-car steel.

Further technical details about these various submodels are given...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Introduction
• Chapter 1 Modelling the Effects of Temperature Sensitive Pressure Relief Devices on Tankers Exposed to Engulfing Fires
• Chapter 2 Practical Engineering for New Fire Protection Problems: An Application to Natural Gas Fueled Bus Facilities
• Chapter 3 Background on Facilities Modification for Natural Gas Fueled Bus Use: Part 2
• Chapter 4 Characteristics of Natural Gas Leaks in Bus Garages
• Chapter 5 Expert Systems Applied to Spacecraft Fire Safety
• Chapter 6 Assessing Community Fire Risk: A Decision Analysis Based Approach
• Chapter 7 Calculation on Axial Forces Generated in Restrained Pin Ended Steel Columns Subjected to High Temperatures
• Chapter 8 A Case Study of Fire Protection in Large Stadia: The San Antonio Alamodome
• Chapter 9 A Systemic Approach to Fire Safety Offshore
• Chapter 10 Salt Water Simulation of the Fire Smoke Movement in an Atrium Building
• Chapter 11 Salt Water Simulation of the Movement Characteristics of Smoke and Induced Air in a Room-Corridor Building
• Contributors
• About the Editor
• Index