Tube velocity limit for air cooled exchanger (API 661)

[engrom]

Looking for some guidance on max tube velocity calculation for wet saturated gas in air cooled exchanger. The exchanger is designed as per API 661, but this standard does not provide any guidance on tube velocity limit. TEMA provides rho-v2 to calculate tube side velocity, but it is explicitly mentioned for liquid service. Can this be used for gas? Is there any other standard which provides a clear guidance on calculating tube side velocity? (Note: this is not for a new cooler, but for cooler in service which has majority of tubes with corrosion within tolerance and some plugged. Due to plugged tubes it will increase the tube velocity. An increased tube velocity may affect corroded tubes, hence the reason to understand the max tube velocity limit)

 

[georgeverghese]

For carbon steel lines, use 25m/sec as the upper limit to avoid erosion corrosion.

 

[engrom]

Thanks georgeverghese. Is there any calculation method or reference to work out tube velocity in such cases?

 

[gerhardl]

Tube velocity: Depending on how exact you want to be, and what and how wet 'wet saturated gas' you are talking about (saturated gas is not wet in itself). Any condensation in the system must be teken care of, and this requires specialist engineering and 'steamtraps' to avoid liquid hammering.

For very rough speed calculations of dry gas: volume per time through the system divided with pipeline area, but it can be done much more complicated and precise, especially if pressure and temperature varies.

 

[georgeverghese]

Gas velocity would be highest at the inlet to the sat gas fin fan cooler. If you know the no of tubes in the inlet first pass, say this is N1.

velocity = mass rate / rho-gas / N1 / (0.785*d^2)

velocity = m/sec, mass rate = kg/sec, rho-gas=gas density at inlet in kg/m3, d=tube id in m.

Since you say some of these tubes are corroded, these must be carbon steel tubes ? Erosional velocity limits > 25m/sec would be permitted for say lean duplex SS (UNS 31803).

 

[EdStainless]

Goerge, generally 2205 is thought of as a 'standard' duplex with the 20%Cr grades as being lean, alloys such as 2101 and 2003.
If there are no chlorides in the system then 304 would stand up to this very well.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[gerhardl]

We are, as usual for forum questions, missing facts for further comments. Both chemical composition, temperature and content of gas and wet fluid, pressures, temperatures and all technical details for the exchanger is necessary to give anything but rough indications.

As an illustration you would use a wide velocity range, depending on details, for saturated water steam
The steam velocities or speeds below are commonly recommended as acceptable for steam distribution systems (courtesy of engineering toolbox). Note that steam and water mix is recommended at less than 25 ms.

The steam velocities or speeds below are commonly recommended as acceptable for steam distribution systems:
Steam System
Velocity
(m/s)
(ft/s)
Saturated Steam - high pressure
25 - 40
82 - 131
Saturated Steam - medium and low pressure
30-40
99 - 131
Saturated Steam at peak load
< 50
< 164
Steam and Water mix
< 25
< 82
Superheated Steam
35 - 100
100 - 300
Saturated steam - low pressure - is common for heating services and secondary process pipes.
Saturated steam - high pressure - is common in powerhouse, boiler and main process lines.
Superheated steam is common in power generation and turbine plants.

 

[Iomcube]

georgeverghese
Although you have given a definite value to upper limit but while dealing with compressible gases an important factor is pressure drop. So b/w inlet & outlet if there is substantial pressure drop the outlet velocities will be higher than inlet. Designer must take this into account as well because with the passage of time fouling will occur & the above phenomenon can't be neglected