how to calculate DETA T (Temperature Range) for thermal shock

[Saver2008]

Hi everybody!!!

I need of your support. I have a pipe that will have SUPERHEATED STEAM. The temperature of this steam is 300 celcius. I`m using carbon steel pipe ASTM A-106 gr B. I need to find out the minimum temperature which a thermal shock can happen. -We are assuming that the time of changing from ? MINIMUM TEMPERATURE TO 300 Celcius is instantly (I mean 1 or 2 seconds). How can I calculate this minimum temperature? Or...is it possible?

Thanks!!!!

 

[BigInch]

From rain? Ambient temp F - Height of the clouds / 3500 will be the temperature of rainwater. Probably not likely that you will come close to that temp, unless you really got a soaking. Actually I think it will depend on what scenario you are considering. Flood? Firewater deluge? Will it be applied initially only to the top half of the pipe? Or the bottom half. Will you get thermal bowing?

I don't think I would use anything less than minimum ambient temperature during the last 24 hours, unless I had a good reason to use something else.

What would you be doing, if you knew that you could not fail?

 

[Saver2008]

Dear BigInch,

First of all thank you very much for answering. In the sceneario that I`m assuming is:
-Within the pipe will bring 0 kg/hr of steam at 300 Celcius, then in 1 or 2 seconds later, the same pipe will have 40 000 kg/hr of superheated steam at 300 Celcius and 15 kg/m2. I`m not considering that the pipe is insulated (in reality it is insulated but for thermal shock temperature range I`m not taking into account). The idea is to know at what minimum temperature does the pipping meet a thermal shock.
Is there a formula or equations that can tell me that? Or what considerations should I take in order to calculate that minimum temperature? I need to find that minimum temperature in order to know that if I have a pipe from X minimum temperature and after 1 or 2 seconds I have 300 celcius of superheated steam inside of the pipe will be produced a THERMAL SHOCK.
What do you think?

 

[TD2K]

I saw your post on another forum where you said this is for a steam assisted flare.

From similar flares I've worked on you have a steam header out to the flare stack, a control valve (or valves) at the base of the flare stack to add steam to the flare stack tip. There is also a series of restriction orifices around the control valves to keep a small flow of steam up to the tip at all times. The continuous flow of steam keeps the steam piping up to the tip hot. The flow rate is usually calculated to avoid condensing any of the steam as it flows up the piping from the control valve at grade to the tips. That is, the heat loss from your piping to ambient air is made up by the drop in superheat in the steam (you gain significant superheat as you drop steam pressure to not much above atmospheric through the restriction orifices).

The main steam header out to the control valves sits under normal steam pressure and effectively no flow. Due to heat loss, the steam cools to saturation temperature. Some of it will also condense so you need proper steam trapping in the pressurized header to remove the condensate. When the control valves open, the steam flows up to the flare tip and the saturated steam supply is replaced by superheated steam. We did not have any thermal shock issues with the piping on our flare stack steam piping.

The piping upstream and downstream of the control valves is designed for superheated steam temperature plus any allowance your specifications require.

The piping does not go to superheated steam temperature instantaneously. The thermal mass of the pipe and to a lessor extent the insulation is heated by the hot steam and the heat transfer coefficient between the steam and the surface of the piping.

 

[BigInch]

So, you know the maximum temperature. You should consider the insulation.

Temperature of the pipe will be increasing according to the heat transfer rate and heat capacity. It won't take long. If the inside of the pipe is substantially different than the outside, that would create a thermal shock. If internal radial expansion exceeds exterior expansion; that's what you're interested in, or some other condition?

What would you be doing, if you knew that you could not fail?

 

[Saver2008]

@TD2K
Thank you very much for your support. You really knew what I´m looking for. I`m going to explain the reason for what I want this and what I preted to do:
-I have a flare stack that will flare or burn Gas with some composition. It will produce SMOKE so the flae will be assisted with steam up to 20% of total flaring. The quantity of steam is 40,000.00 kg/hr. I have my pipe of 8"Ø sch 80 with a insulation of 2" thickness. The total length of 8" piping is almost 120 mts including loops and so on. My client is asking me for a MINIMUM QUANTITY OF STEAM ON PIPING IN ORDER TO PREVENT OR AVOID THERMAL SHOCK BUT HE IS ASKING ME FOR THE CALCULATION. In order to solve this I was thinking in the following:
--I have the flow of steam for maximum scenario, including temperature and pressure, even enthalpy and entropy and specific heat can be calculated or obtained. Thermal shock is a SUDDENLY CHANGE OF TEMPERATURE ON MATERIAL SO IT MAY CRACK. So, I was thinking to obtain the minimum temperature which thermal shock can be produced. In any case a formula for minimum temperature does not exist, I was thinking to make a stress analysis with thermal conditions from 0 celcius to 300 celcius in order to know the stresses which pipping can be able to resist without problem. Having the minimum temperature (it will be according to results from thermal analisys) I`m going to be able to have the range of temeperature which a thermal shcok can be produced. Then using a formula for calculating the HEATING ( q = m cp dt/t ) and ( ms = q/he) and having the temperatures (one will be 0-300 celcius and the other from 0 to minimum temperature in celcius) and with that I can have the Heat for minimum temperature so I`m gonna be able to calculate the MINIMUM QUANTITY OF STEAM.

Accordint to your writing, I can see that THERMAL SHOCK is no produced on steams lines for flare stack issures.....but my client thinks that this effect can be prodcued when on pipping exists suddenly different temperatures conditions......what do you think about it.

@BigInch,
I do not consider the insulation because the temperature for thermal shock is the temperature on my piping, not on the outside layer of my insulation. So in order to NOT TO CONSIDER MORE ELMEENTS THAT MAKE MY EQUATION LARGER, I think I can reduce it or limit it to just piping temperature and work with that. I think insulation purpuse is to not to have any heat loss from origin of steam to the end in which the steam will be used (n my case to the top of my flare) but I do not care what is the heat loss because I have the insulation and its thickness by original specs....I`m just thinking about the temperature (internal temperature of piping)so a thermal shock can be produced....what do you think?

 

[BigInch]

With the overall heat transfer coefficient of the pipe and insulation, you will be able to find the differential temperature between inside the pipe and atmosphere. If you include the insulation, the diff temperature across the pipe will remain smaller. Assuming your steam flow is long enough to reach thermal equilibrium, the heat capacity of both the pipe and insulation drops out of the equation, leaving only heat conductivity and you arrive at a steady thermal state with internal temperature equal to steam temperature. It is then only remaining to find the temperature of the pipe's outer surface inside the insulation. If you do not include the insulation, you may have to use ambient temperature as the temperature of the outside of the pipe.

What would you be doing, if you knew that you could not fail?

 

[TD2K]

I don't know how to calculate what you are asking for that to be able to convince your client or even if such a calculation exists.

As I said, your flare line up to the control valves is going to sit full of essentially saturated steam under only a small flow rate around the control valves up to the assist steam tips. Upstream of the control valve, the piping will be about 200 deg C, going to 300C is not uncommon in any of the steam assist flares I've seen. Downstream of the control valves due to the small amount of steam you bypass around the control valve, that piping will be 100+C, again, I don't see a 200C change 'thermally shocking' the system though your piping has to allow for the resulting thermal growth.

I've also seen steam turbines with autostarts where the inlet steam on/off valve pops open say due to a loss of flow. The steam turbine being rotating equipment is a lot more sensitive to thermal shock than steam piping. In this case, the steam turbine sits full of exhaust steam at its saturation temperature and the steam leaving the turbine blades for smaller backpressure turbines isn't a lot cooler than the supply steam temperature.

I talked to one of our stress engineers and he wasn't even sure where to begin 'quantifying' the case you've raised nor had he ever seen it raised an a question/concern. As part of our design we look at the thermal expansion of the overall header from cold to operating temperatures but not some "temperature rise over x seconds".

If you have someone familar with the piping codes you are designing the steam system to I would ask them if this issue is discussed in the codes.

 

[BigInch]

That's because people usually only look at the steady state problem, if they look at all. I also doubt there will be any problem, however if the client wants to consider the thermodynamic aspects of the problem, it certainly can be done; simply include the heat transfer rate and heat capacities of the materials involved, do a finite difference analysis of the interior to exterior of the pipe and insulation and step through until you reach the thermal steady state. Check the expansion stress at each step. It's ususally more productive just to sit down for a few hours and come up with the answer, rather than to spend days arguing with the client about why you don't have to and resolve nothing.

What would you be doing, if you knew that you could not fail?