Mininum gas velocity for glycol injection 4

[fdomin]

Which is the minimum gas velocity required in natural gas exchangers (gas-gas/chiller) to provide adequate glycol distribution and good hydrate inhibition?
Is it velocity the correct parameter or others may be used
(e.g. ro.v2)?

Thanks

 

[whoislee]

Pardon me why is glycol injection required?

Normally, natural gas is dehydrated at the production site (offshore or onshore) using TEG dehydration prior to sending it to downstream equipments for further processing.

Methanol injection is done to protect the piping from the gas well to the gas processing facilities. Glycol is seldom used or heard of for this function.

Perhaps, you may want to consider building a gas dehydration facility to remove most of the water prior to sending it for further processing.

There are 2 sides of a coin
One is to give, one is to take
Give until it hurts with a smile

 

[TD2K]

Glycol injection is done onto the face of chillers that are taking the gas below hydrate temperature in order to recover heavy hydrocarbons. Downstream of the chiller, a three phase separator is used to separate the gas, HC liquid and glycol. In this service, ethylene glycol is commonly used.

Adequate glycol injection is usually accomplished by ensuring you have proper dP across the glycol spray nozzles and the design of the spray nozzles themselves to distribute the glycol across the face of the exchanger inlet tubesheet. Gas velocity is not usually intended to redistribute poor glycol distribution.

The gas flow and the temperature drop across the exchanger will give you how much water will condense. Simulators or the Hammerschmidt formula will give you the required 'rich' glycol concentration coming out of the exchanger. A mass balance around the exchanger on the glycol side will give you the required flow rate based on the lean glycol concentration your regeneration system will give you (typically about 80% for this type of a system I believe). A significant safety factor is commonly applied to make sure you have enough glycol to prevent hydrates.

 

[fdomin]

TD2K,
I agree with you that glycol distribution is acomplished by good Dp in spray nozzles and correct spray design (e.g spray angle)
I want to know if a minimum gas velocity/ro.v2 inside the tubes is required to avoid glycol settling (in the tubes).
What about a maximum gas velocity/ro.v2?

thanks

 

[TD2K]

I have not seen that criteria for glycol spray systems. Perhaps someone with more experience has some information they will share.

 

[maddocks]

Typical velocities are a bit of a loose criteria, but we have used 12-15 ft/sec as an acceptable minimum. Having said that, we're not overly concerned if the velocity is too low.

Jim Maddocks, P.Eng.
jmaddocks@gasliquids.com

 

[Mike6158]

Personally I have found that designing for a proper spray pattern and proper flow rate of EG are the two most important parameters to pay attention to in the design of an EG injection system. A lot of designers will over circulate glycol in the incorrect belief that this will improve dehydration. In most cases over circulation will result in increased hydrocarbon entrainment in the rich stream. This typically overloads the reboiler.

I developed an Excel spreadsheet for designing EG injection systems. I used it to redesign the EG injection systems in 4 of our facilities with 100% success. We successfully lowered EG injection rates from 5-10gpm to 2-3gpm.

 

[jcaiken]

My understanding is that the object of the glycol injection was to enable the glycol to wet all the metal areas likely to be cold enough to condense water vapour from the gas. The water droplets would then dissolve into the glycol rather than wet the metal surface and cool further to form hydrates. The effect of the surface film distribution on the overall performance is as important as the initial spray distribution.

As a result the velocity should be high enough to ensure a good, even distribution of liquid onto and down the metal surfaces and low enough to prevent the glycol film from remaining in the gas phase or being re-entrained.

 

[Hookem]

It is not the velocity of the gas stream that is important; rather, it is the spray pattern. One must strive towards atomizing the injected fluid. The smaller the particles, the more the surface area for contacting with the gas stream. Virtually any nozzle spray company with a technical sales/engineering staff can assist. I have used Texas Spray Systems, but others too can be used.

 

[Mike6158]

Glycol injection is a "contact sport"

And here's a few rules of thumb...

(1) Try to put about 80% of the glycol that you inject at the front end of the exchanger train. If there is a front end split, divide the amount injected evenly between each exchanger.

(2) When you select a nozzle try to get a very finely atomized pattern. Spend some time and calculate the distance the nozzle needs to be from the tube sheet so that the tube sheet is fully covered by the spray pattern. Try for about 100 psid (the nozzle mfg shuld be able to give you cone angle and should want to know what the nozzle delta P will be). Don't overlook downstream pressure drop in the delta P calc. By that I mean exchanger pressure drop. If you have exchangers in series the downstream exchanger will operate at some pressure that is lower than the upstream exchanger. Simple yes but easily overlooked.

(3) When you install the nozzle try to install it in such a way that it is easily removed. Nozzles wear out. Also, install a block valve before the nozzle piping enters the vessel and then a vent valve downstream of the block. If the nozzle plugs up you can sometimes shut it in and open the vent valve to allow gas to flow backward and unplug the nozzle. Sometimes that works and sometimes it doesn't.

(4) Pay attention the the rich return stream composition. Specifically it's freeze point. Try optimize the lean injection rate to result in a return stream that is at the very lowest point of the curve.

(5) Don't over circulate!

(6) Pay attention to the residence time of the 3-phase separator boot. I like to have no less than 45 minutes. Doing so will minimize the amount of hydrocarbon entrained in the return steam. This will cut down on reboiler load and stack emissions to some degree.

(7) Don't over Cicrculate!

 

[Mike6158]

The velocity of the gas stream is not unimportant either. It and the viscosity of the return stream are very important things to consider. You don't want all of your injected glycol stacking up in the equipment and then "returning all at once". If the velocity is too low there will be problems.

Years ago I worked for an E&C firm. A customer had purchased some heat exchangers and wanted us to build a gas plant with them (a long time before PSM). The design inlet gas rate of the gas plant was 15 MMscfd. The exchangers came out of a 50 MMscfd plant. We tried to get the customer to at least let us plug some tubes in the exchanger (to increase tube side velocity) but they would not allow it. To keep the story short, we had 2 MMscfd for startup, it was a 950 psig design and we had 700 psig. Glycol stacked in the system, and freezing problems occured. The customer drilled the nozzles out and tripled the glycol injection rate which worsened the problems. Eventually the gas rate came up and we put new nozzles in. Once that happened the problems all but went away. Years later I came to work for that company (still here) and completely revamped the glycol injection system in 5 of our plants. The first thing I did was to cut the circ rate drasticly and do a rigorous analysis of the injection nozzles and their installation. Glycol losses dropped significantly. Freeze-ups stopped occuring. The reboiler firing rate dropped significantly. All in all things were better.