Steam Hammer?

[KevinNZ]

Can a sudden change Steam flow cause a pressure change like water hammer does?

I have a long (2km) large dia (1050mm) steam main. (Flow is 110 kg/s @ 4.5 bar.g.) We have a vent to control pressure at the upstream vent. If the flow is stopped at the downstream end and directed to the vent does the change in momentum cause a pressure change at the downstream end?

Does anyone have refernce that covers this?

Thanks

 

[davefitz]

Yes , steam hammer can cause similar pipe transients as a waterhammer event.

In large fossil steam boilers, the HP main steam line is at 1000F at 2500 psig, approximately 16" inside daimeter made of P91 or P22 material, about 200 ft long from boiler to turbine inlet. When the turbine strips, the stop valves close in 0.2 seconds, causing the all of the stored mass in the steam line to change veleocity from 400 fps to 0 fps in 0.2 seconds. If you compute the forces caused by this deceleration, they are significant and there is also a resulting pressure pulse generated.

In general, for the valves which are permitted to close at a slower rate ( ie , bypass valves), the rate of closure in the valve positon range 20-0% open is restricted to limit the magnitude of the transient.

 

[zeven]

The Max pressure rise can be calculated by multiplying the density of the fluid by speed of sound of fluid by velocity of fluid. You can then estimate the forces by pressure X area
This value will only be achieved if the valve closure times are less than the time it takes for a sound wave to travel from the valve(which is closing) back up to the reservoir and return - in your case a distance of 4km.
The speed of sound can be reduced by the elasicity of the pipe wall, thus reducing the pressure rise.
Max load on any single straight length of pipe can be calculated by factoring the max pressure rise by the a value of leg length divided by the wavelength of the transient (valve closure time x speed of sound). This assumes linear closure and bends of 90 degrees at both ends.

Alternativly you could buy a fluid transient program and input results into pipe stress analysis program and then perform a time history analysis. Wavenet was one from the 80 & 90,s

 

[fvincent]

I would add that normally the main reason to have such header pressure elevation is the time between the instant when valve closure happens and pressure control action. You could model the problem by as a resistive capacitive system (dividing your entire system in sub-volums and connecting then with pressure loss coefficients).
Then you begin to close the valve and detect the pressure elevation in the pressure element point. Given a PI control function and a maximum velocity for valve actuation you can have the RC response. Then add the max value calculated as ZEVEN has showed.
This will be a good answer for the pressure evolution along time.
Hope it works for your case
regards



fvincent

 

[KevinNZ]

I my case the calculation is going to be even harder because the steam is saturated and when the pressure rises the steam will condense!

Cheers
Kevin