Hydrogen Tank Leakage in FDS

[Gesa Pottbrock]

Hi all,

I am learning how to use FDS at the moment.

I want to simulate a hydrogen tank that is leaking.
I am planning on defining a closed room in my FDS model and using localized leaks via HVAC to model the leak of the tank.

I am wondering now, what the smartest way is to fill the enclosed room with hydrogen up to a certain pressure. The only way I know is to define a vent on the bottom of the tank and let hydrogen flow into the tank (similar to 12.1.1).

But I don't want to simulate the inflow of the hydrogen. Because the room is an enclosed space, it will have a different pressure zone than the room the tank is positioned in.

Is there any way to define the enclosed room as "This Zone is filled with hydrogen at 500 bar pressure"?

Thank you in advance for your answer.

Best Regards,

Gesa

[fde]

If your objective is the assessment of the conditions in the room, then you do not need to model the hydrogen tank in 500 bar. If you know the leakage rate, then you can impose this boundary condition to a VENT as a supply. 

[Gesa Pottbrock]

Thanks for your answer.

Yes, I want to assess the conditions in the room. I have not received all the parameters for the tank I should simulate, so I am sure yet if I know the leakage rate. I am brainstorming the possibilities I have for describing the Problem via FDS.

Do you know if there would be another way of modeling that problem as I described above?

[Kevin McGrattan]

If 500 bar is an important consideration, FDS is not the right tool for the job. FDS is not appropriate for pressure vessels.

[dr_jfloyd]

fde's advice is good

FDS uses the Ideal Gas Law as the equation of state. The Ideal Gas Law is not a good approximation of gas behavior at high pressure. At 500 bar you will probably have > 10 % error in the tank conditions. I would not use FDS to model the inside of the tank.

FDS is also a low Mach number code (Mach number limit of ~0.3). At a tank pressure of 500 bar you are well above the critical pressure ratio for choked flow and FDS would not predict the correct leak rate.

Your best bet is to do a hand calculation based on your leak area ato determine the time-dependent leak rate of the tank and impose that as a boundary condition keeping in mind that you may need to make adjustments to how you define the leak to keep a low Mach number.

[Gesa Pottbrock]

Thank you all for your answers, they were very helpful.

Good to know that the Ideal Gas law is used, which is indeed not super accurate at describing the behavior of real gas.
Now I can assess better what is realistic/ not realistic to simulate with FDS.

If I did my research correctly hydrogen tanks usually operate around 500-700 bar.

Have a nice weekend :)

Gesa

[Gesa Pottbrock]

Hello everyone,
I have tried to figure out, how fast the hydrogen will escape out of the tank. I found a paper, which states equations to calculate the leakage velocity on hydrogen tanks.
The paper is called "A numerical study of hydrogen leakage and diffusion in a hydrogen refueling station"( https://doi.org/10.1016/j.ijhydene.2020.03.140 ) and I used the equations 8-11 with the given quantities. I have attached a screenshot of the equations below.

I have calculated the leakage velocity to be 1332 m/s which seems way too much. The paper simulates the leakage of a hydrogen tank with ANSYS and in Fig. 7a one can see the leakage at different times. There the leakage travels roughly 3m in 0.25 sec which would translate to 12 m/s which is a totally different value.

Does anyone have experience with leakage velocities and could the value 1332 m/s be realistic?

Best Regards,
Gesa

[Gesa Pottbrock]

I should have clarified:

-b is the co-volume with b=7.69 *10^-3 m^3/kg

-\kappa = 1.4

-R_{H2}=4124.24 J/(kg K) is the real gas constant for H2

-p1 is 40 MPa

-T1=T3=298 K

[dr_jfloyd]

500 bar to 1 bar will be choked flow (the critical pressure ratio is ~2). The flow speed will be Mach 1 in the throat of the leak which for H2 is ~1300 m/s.  As the jet entrains air outside the leak the speed will slow. For a pinhole leak (a few mm diameter) the speed would be greatly reduced at 3 m (1000 times the leak diameter) and having an average speed of 12 m/s over that distance could be seen.

[Gesa Pottbrock]

Thank you for your answer.

Oh yeah, that makes total sense with the jet entraining the air and therefore the reduction of flow velocity.

So, if I supposedly want to model the leakage with vents in FDS do I get a Problem due to the flow speed being 1 Mach right at the leak or could I just choose the flow rate to be 12 m/s?

[dr_jfloyd]

FDS is a low Mach number code. This is the first item noted in the overview of FDS in the Technical Refernece Guide.  

FDS Technical Reference Guide

Chapter 2

Overview of the FDS Model

This chapter presents the governing equations of FDS and an outline of the general solution procedure.
Details are included in subsequent chapters. The purpose of this chapter is to highlight aspects of the
solution methodology that make it practical for thermally-driven flow simulations, in particular fire. Some
of the major features of the model, in its default operation, are:

• Low Mach, large-eddy simulation (LES)

If you want to model the hydrogen jet, you need to, as best you can, preserve the jet momentum.  At the leak you have 100 % pure hydrogen flowing at sonic speeds.  There a number of approaches in the literature for predicting the entrainment of jets as a function of distance from the leak.  At some distance enough air will be entrained to reduce the speed to being "low-speed" (Mach <~0.3); however, at that point you no longer have pure hydrogen you have a mixture of hydrogen and air.  You will need to develop an approach for doing this.