Frequently Asked Questions

Questions related to use of VessFire

What information do I need to set up a case in VessFire?

  • P&ID covering the segment
  • Layout of segment
  • Length of pipes
  • Pipe specification for the relevant pipes in the segment (diameter, wall thickness, material type, corrosion allowance, mill tolerance)
  • Phase type for all pipes. Pipes are taken to be either fully liquid or fully gas
  • Ultimate tensile stress (UTS) for material in pipe and vessels
  • Diameter, wall thickness and length of vessel
  • Liquid level in vessels
  • Composition entering the segment (inventory)
  • Temperature and pressure of inventory (usually operational conditions).
  • Heat exposing scenario (heat load, time duration)
  • Ambient temperature
  • Information about blowdown valve (BDV) (activation time including registration of fire), pressure safety valves (PSV) (set pressure, full-opening pressure and reset pressure) and upper limit for the release rate

How should “longitudinal position” and “angular position” be set? Can we define a blowdown valve at the bottom of a vertical or horizontal vessel?

If you have a horizontal vessel, you can change the angular position from 0 to 180 degrees. 0 degrees at the top of the vessel, and 180 at the bottom. If you have a vertical vessel, you let the longitudinal position be 0 (bottom) and the BDV, or PSV if you like, will be placed at the bottom.

longitudinal position

How is dense phase (supercritical fluid) reported in VessFire?

In VessFire 1.2 a dense phase is reported as “gas mass”.

BDV time delay - what about valve opening time?

BDV time delay is defined in VessFire as delay plus opening time. It should be considered as “time till fully open”. Within that time, the release rate is zero, which is a conservative estimation. BDVs normally open at a rate of 1 second per inch.

For which temperature corresponds the UTS value used as input?

The initial input value should be the one at 20 degrees Celsius. The UTS curve is governed by the chosen material. So, if you use the initial values given in the Scandpower guidelines for the corresponding materials (UTS curves) you should be pretty safe. Of course, if you have UTS data for the materials, you can make your own curves and use the proper initial values. However, from experience, such data are hard to obtain.

Can I define the background heat load to a limited area or length of vessel?

The background load covers by default the entire vessel, apart from the area covered by the peak heat load. The best option would be to lower the overall background load. The average fire load is not meant to represent one particular fire, but the effect of a range of fire situations that can be possible. The first is a question which would extent the scope – can VessFire deal with low temperature effects in vessels during blow down? VessFire has been used for low temperature analyses for a long time, both for vessel and pipes. It has also been tested against cold blowdown experiments. There will be a heat transfer coefficient on the outside of the vessel and pipes to take care of the heat transfer in addition to radiation. It is the same physical phenomena during a cold blowdown as for the heat exposed case. There will also be low temperature profile for the pipes. What VessFire does not calculate, is a precise temperature around the exit of the vessel and the temperature in the blowdown pipe. This area is strongly influenced by the exit velocity and need to be studied in detail. We have a special tool to investigate these issues.

Regarding valves. What is the contraction factor parameter?

The contraction factor (also known as discharge coefficient) is an area ratio between the effective flow area and the physical flow area.

How do I specify the Blowdown line if I have several process equipment units in a segment?

The blowdown line is modelled to include the pressure drop from the segment to the blowdown line downstream. When setting up a fire scenario depressurization case, it is usually advisable to proceed with the most conservative approach. If a blowdown line has varying diameters, applying the smaller diameter for the entire stretch could be pertinent. For cases with multiple process equipment units, the blowdown line can be best approximated as the stretch of pipe lines that causes the largest pressure drop between the segment and downstream of blowdown valve.

What is the "blowdown line"?

Screenshot_1

The blowdown line is the length of pipe between vessel and flow orifice.

The input felds is located under “Vessel” in the “BDV” tab. These fields are used for calculating pressure drop from vessel to flow orifice. Please note that the pipe used as blowdown line should still be included in the “Pipes” group.

There are multiple ambient temperatures. How are they used in the simulation?

Ambient temperature fields are located in the following location (click to enlarge):

under Vessel in the Condition tab

under Vessel in the Condition tab


for pipe heat loads

for pipe heat loads

To understand how the ambient temperature is used, it important to note that it is related to heat loads and not vessels or pipes. The ambient temperature is used to estimate a flame temperature. Consult the documentation here for further information on flame temperature.

Using different ambient temperatures for pipes and for vessel may be applicable in some circumstances, for instance when the pipes and the vessel are situated in separate fire areas. In general, however, the amibent temperature should be the same for all pipes and vessels in the same segment.

Can VessFire deal with low temperature effects in vessels during blow down?

VessFire has been used for low temperature analyses for a long time, both for vessel and pipes. It has also been tested against cold blowdown experiments. There will be a heat transfer coefficient on the outside of the vessel and pipes to take care of the heat transfer in addition to radiation. It is the same physical phenomena during a cold blowdown as for the heat exposed case. There will also be low temperature profile for the pipes. Nevertheless, a precise temperature around the exit of the vessel and the temperature in the blowdown pipe is beyond the scope of the calculations performed by VessFire 1.2. This area is strongly influenced by the exit velocity and need to be studied in detail. Such kind of analysis are done by the use of VessFire 2.0

Technical Questions

What is the calculation sequence for a typical VessFire algorithm?

  1. Geometry set up and material property collection.
  2. Initiation of calculation.
  3. Flash calculations to find composition in liquid and gas as well as evaporation, condensation and phase properties.
  4. Release rate calculations.
  5. Calculation of liquid height, pressure and density.
  6. Calculation of heat transfer coefficients.
  7. Calculation of inventory governing equations for liquid and gas.
  8. Calculation of external heat load.
  9. Calculation of temperature distribution in vessel shell considering external heat load and inventory temperature.
  10. Calculation of shell stress and comparison to the acceptance criteria for the elevated temperature.
  11. Next time step starting at step 3 and repeating all steps up to 11.

Why is water and hydrocarbon level listed as inputs? Are not inlet composition, temperature and pressure sufficient to calculate the different phase volumes?

The water, HC liquid and gas volume fractions could indeed be obtained from these initial parameters. However: Small deviations in composition would imply large deviations in volume fractions, and therefore also in total mass for the segment. The inlet composition is not necessarily the same as the global segment composition. Liquid and water levels are widely measured and serve as more consistent parameters for calculating phase volumes than the composition. In conclusion, VessFire finds the phase compositions from flash-calculations, but not the size of each phase.

Which equation of state does VessFire 1.2 use?

The Peng-Robinson equation of state is used to determine the phase equilibrium. VessFire 1.2 use SUPERTRAPP for flash calculations. A P-H flash (pressure and enthalpy is known) is performed for calculation of mass balance between phases, composition for each phase, temperature and properties like density, viscosity, conductivity, compressibility, etc. To find pressure, a T-D flash (temperature and density is known) is used.

Are thermal stresses due to temperature gradients included?

If the temperature gradient should give high stresses, the material would yield and the stress would not be above yielding stress. As VessFire are testing for UTS, the yielding stress is not of large interest. The temperature gradients are not able to rupture a flexible shell. As the temperature increase, the yielding stress will also be lower.

Is the pressure uniform throughout the segment in VessFire simulations?

The pressure is equal for all components (vessel and pipes). Nevertheless, each pipe has a local temperature and a local heat transfer is calculated.

How should flame length affect the heat load input? Does the Scandpower Guidelines hold for bigger fire scenarios?

The principle of applying a peak load and a background load is based on the idea that the peak load shall weaken the steal and the background load shall contribute to the heating of the inventory of the segment. When we implement a segment consisting of vessel and pipes in VessFire, each object are tested by applying these two loads. If a huge fire resides in a module or a smaller jet fire hits an object, the effect on the steel will be the same with respect to strength. 350 kW/m2 is suggested to be a “typical” peak heat load and will most probably occur in a limited area within a flame, even in big flames. The load is for that reason in principle meant to be independent of size. The size of the peak heat load area is of minor importance to the result as long as there is a minimum heated area. The background heat load is more influenced by the size because it is meant to be the average load caused by the energy released in the fire area. There is another effect entering as well: If a huge release occurs, oxygen will be the limiting resource in enclosed areas (or even in partly enclosed areas) and in that manner limit the heat release. The consequence may be that the flame moves outside the module or only a part of the released hydrocarbon will burn.

Flame length as measure of heat impact is not a good measure. Unless iso-curves of heat flux are specified, the length is of no use with respect to evaluation of heat load to the process equipment. A heat load of 100 kW/m2 is quite high as an average. Flame length is often used as basis for risk evaluation, and can of course be used to identify possible interaction points on equipment, based on the distance from the release point. But without giving heat flux as function of location, the term “flame length” is undefined and gives no basis for estimation of heat load.

What is the relationship between peak heat load, global heat load, inventory temperature and steel temperature?

Both the peak load and the global load are influencing the inventory. The steel maximum temperature is taken from the point where the temperature is highest, independent of where it is located. If you increase the global heat load you will increase the inventory temperature as well. This will influence the peak heat load as the cooling effect from the inventory will be less. VessFire take all these effects into the calculation. It is expected that the rupture time will diminish as you increase the global heat load. It is also expected that the global heat load will influence the steel temperature even in the peak heat area.

What is the relationship between water level, HC level and the composition?

The composition must always be given as the actual feed into the vessel/separator. The water and HC levels are the levels indicated on P&IDs. If a zero water level is given, VessFire will take the HC level as the total liquid level and calculate the water portion based on the composition. Ideally, the feed composition and both a water level and a HC level are given; the composition should not be recalculated to adjust to a specific water level.

How is water treated in the VessFire calculations?

A mixture that includes water is treated by separating water from the mixture and performing flash calculations separately. The thermodynamic of water is quite complicated and the separation is done in order to be able to treat higher content of water in mixtures. Mixing between fluids and water is assumed a mechanical mixture. That means no molecular forces are considered between components of the fluid and water. As the mixing is assumed complete, water and fluid will have the same temperature. This is enforced by heat transfer between fluid and water in the energy equation.

What is "BDV line after orifice"?

The temperature in the pipeline downstream of the BDV flow orifice is of importance when testing for brittle fracture after a cold blowdown. This temperature can be found in the “Steel temperatures” graph after, and while, running a simulation.

This temperature is taken some distance after the flow orifice, for a pipe with same dimensions as the blowdown line. The pipe material equals the vessel material.

Troubleshooting

I’m getting the error message "No valid scenario file available in directory"

Check your file path. If you have the period symbol in your path, remove it and try to run the case again.

I’m specifying a HC level, but the simulation only gives a gas (and/water) phase. Why is there no liquid phase?

Specifying a liquid HC level will result in a liquid phase only if the VessFire’s thermo-package finds a liquid phase. If the flash-calculations yield one phase only, then the specified liquid height is not taken into account.

I’m experiencing that pipe length has effect on rupture time, especially at small pipe lengths. What causes this?

All pipes are subjected to a peak load covering 20cm (7.87”) of the length, regardless of other parameters. When using small pipe lengths, this will have an effect on rupture times.

Does the jet flame contribute to inventory heat up? If so, in a segment with 10 pipes, will the jet flame cover 20cm*10 = 2m of pipe length that will provide an extra pressure build-up?

The jet flame does indeed contribute to heating and pressure build-up for the inventory. VessFire handles jet flame in a conservative manner, by including heat transfer from the jet-exposed steel to the inventory. If a less conservative scenario is desired in which fewer pipes are subjected to peak load, this can be handled by combining pipes that has similar specifications to a single length of pipe. This will result in a less conservative pressure profile. However, this is not always recommended for small pipe lengths (especially gas filled) because these are usually weak points in terms of rupture, and this will not be reflected if they are combined as single length of pipe.

Is the water level calculated internally if water level is set to 0 m? For what reason would a volume of water in the vessel not give a mass of water in the segment masses graph?

When the water content is small, the water is treated together with the hydrocarbon. This is acceptable and quite common to do. The Peng-Robinson equation of state does not behave well for water, but for low content it is acceptable. In these situations the water is not identified as a separate phase. That is the reason why you can’t see the water in the results even though it is treated. When there is a water level, we need to treat water separately. Then the water should show up in the results. Vessels that have only oil and gas, and where the water is not regulated, should be modeled without using a water height. This is because the water will most probably be drained with the oil. In that case the oil content comes as a result from the flash calculation. If there is a separate water regulating system where the water height is set, a water height should be set. In this case the water content in the vessel will no longer be a result of the in-flow composition The flash calculation gives the composition of the gas and liquid and the properties for each phase. The flash calculation finds the initial enthalpy for use in the energy equations, the molecular weight for the liquid and gas phase and other properties for the mixture. The mass of liquid is found based on the vessel dimension, the liquid height and the liquid density. VessFire uses two software packages for calculation of thermodynamic and transport properties, one for fluid mixtures and one for water and steam.

Why is the inventory composition different from the inlet composition?

The vessel inventory composition is a result of a flash calculation based on the inlet composition. The inlet composition is the total composition for the feed stream. For that reason, liquid and gas inside the vessel has in general a different composition compared to the inlet composition defined by the user. Nevertheless, if the flash calculation results in one phase only, the vessel inventory composition will be set equal to the initial inlet composition.

The mass balance of oil and gas seems to be wrong, according to my own predictions. Why?

The mass balance shows the mass of the total system, including pipes. If a pipe is entered with phase set to gas, it will not be gas-filled if the conditions actually yield a liquid phase, and vice versa. Thus, the total mass balance may be greater or smaller than what was predicted without considering the actual phase.