5 replies to this topic

[Lucas]

Please, consider the following about a glycol dehydration unit:

Process data:
Contactor @ 104 F, 2500 psig
Gas flow - 85 MMSCFD
Glycol flow - 450 gal/h

Karl-fischer results: (confirmed by chromatography)
Lean Glycol - 99,4 to 99,7 %w TEG
Rich Glycol - 92,7 to 95,6 %w TEG

Glycol losses:
100 gal/day (carried by the dry gas)

Water content requirements:
Dew-point 5 F or 1,5 lb/MMSCF (McKetta and Wehe chart for this conversion)

Water content measurements (ASTM D-1142 with Chandler Chanscope):
Dew-point 63 F or 12 lb/MMSCF

Others variables Ok, level controls Ok, glycol regeneration Ok.

Column Data:
Internal diameter 1000 mm
Packing height 5000 mm
Sulzer Structured

So: 1 - Dehydrated gas is, by measurements, out of spec (main problem); and 2 - Glycol losses are considered high (?);

After some research, I found that there is not much useful information regarding dehydration at this pressure (> 1500 psig). So this leaves me with the following "tough" (I think) questions:

- What is the reliability of the above water content measurements, considering the given process data, the ASTM D-1142 method and also the fact that the gas contains lots of carried glycol? (carried glycol is “said to be removed” by filtration from sampled gas during measurement)

- Since I know there’s a considerable amount of carried glycol at the dehydrated gas, I think Mcketta and Wehe chart is not to be used, so:

- What thermodynamics package is actually reliable in predicting this system at this pressure? Does the Twu models of PROSIM, for example, are? And Hysys new Glycol Packages?

- What about the phase behavior of the system gas-water-glycol at such pressure? Is it possible to extrapolate the behavior at low pressures (< 1500 psig)?

- What is your experience about glycol losses at 2500 psig? Is it true that glycol gas solubility increases unexpectedly with pressure? If yes, does it explain this amount of loss? If no, does it means that the contactor internals are with problems?


Thanks,

Lucas
Process Engineer
Oil & Gas - Offshore Surface Instalations

Experience
Oil & Gas – 2 years
Chemical Industry – 2 years

[Art Montemayor]

Lucas:

I’ve recently had a design experience with a TEG dehydration unit close to your unit’s capacity.

My unit had the following characteristics:

Gas Capacity = 70 MM Scfd at 1,775 psig & 110 ^oF;
Lean TEG into Contactor = 99.85% wt.
Rich TEG out of Contactor = 96.37% wt.
TEG losses should be approx. 10 – 12 gal/day
Stripping gas flow rate = 7 to 10 Scf/gal TEG
TEG flow rate = 480 gal/hr
Product Gas Dew point = -7.0 ^oF

I would expect your TEG flow rate to be = 590 gal/hr (not the 450 gal/hr you report) and your stripping gas to be in accordance with the above estimate. You have not mentioned stripping gas. If you don’t have stripping gas applied correctly, in my opinion you will not obtain your expected dew point. You must strip your Lean TEG down to 99.8 % (or better) in order to have a chance at getting your product gas dew point lower.

You are losing far too much TEG. You are probably entraining it out the top of the contactor. For this high pressure natural gas (I’m assuming 85+% methane (volume), I would not have opted for TEG dehydration. I would have preferred an adsorption unit since the water content is so much lower with 2,500 psig gas than with 1,800 psig gas.

Yes, you can expect the amount of dissolved gas in the Rich TEG to go up with the increased gas pressure. This will show up in the TEG flash drum.

I hope this data and comments help align your unit out.

[joerd]

Re. your simulation questions - I have been told that the implementation of the Twu method in Hysys is wrong, so better wait a little until they patch it. Also, its Glycol package is infamous for its inaccuracy. So Hysys is not going to be much help, even if the results agree with your measurement it might be a coincidence. I would try to call BR&E (Promax) or the Virtual Materials Group (VMGSim) and see if they can help you out.

[Lucas]

Art, thanks for your answers

Regarding TEG and stripping gas flow rates, that’s it, right now they’re below recommended conditions, although it does not justifies dew-point measurements around 65 F.

About composition: gas from dehydration has ~ 71 % v/v CH4, > 0.5 % v/v C6+, MW ~ 23.5. This is compressed gas for gas-lift offshore oil production.

About “glycol gas solubility”: I meant Lean TEG solubility in the gas phase – TEG in gas – is it true (or do you believe) that it raises unexpectedly with pressure? (so it would be normal to have high glycol losses at high pressures)

Now I’m really concerned with:

- Pressure: in this case, 2500 psig, and from other units starting now - 2850 psig, 300 MMSCFD. “Normal designed” units will perform well at these conditions? Is it normal to design TEG dehydration for that pressures?

- TEG losses: is column design for high pressures aware of it? In this case, those 1000 mm of internal diameter are sufficient for liquid separation? TEG losses around 1.2 gal/MMSCF are “normal” in six different units that I know (similar operating conditions)…

- Dew-point measurement: does cooled mirror technique works? At high pressure? With entrained glycol and heavy condensing hydrocarbons?

Thanks again,


Lucas

[Lucas]

Joerd

Thanks. Right now I don't have PROMAX license, but I'll contact BR&E.


Lucas

[IsaChem]

Just in case you could be interested here is a page about Natural Gas Dehydration Unit With TEG /
Process modeling and simulation with ProSimPlus.