Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

- - - - -

Downhole Choke Hydrates

downhole choke hydrate

This topic has been archived. This means that you cannot reply to this topic.
7 replies to this topic
Share this topic:
| More

#1 ChemEng01

ChemEng01

    Gold Member

  • Members
  • 84 posts

Posted 10 February 2015 - 08:22 PM

Hey Guys

 

Has anyone ever had experience with or modelled a downhole choke for a gas well to manage tophole pressure and reduce temperature drop across wellstream choke?

 

I'm looking at a typical gas well and hydrates / wax downstream of the wellhead choke. The flowing tubing head pressure (FTHP) from the well could range from 90 - 160 barg at flowing tophole temp approx. 35C and the pressure is let down across the choke to 75 barg.  

 

At an FTHP higher than 100 barg methanol & PPD requirements are excessive.

 

Apparently when the pressure is reduced downhole the gas stream will heat up because of the high ground temperature. This results in a lower flowing tubing head pressure (FTHP) and a higher temperature downstream of the wellhead choke valve.

 

So the plan is to install a downhole choke to limit the FTHP to 100 barg at FTHPs higher than this.

 

I have simulated this in HYSYS using pipe segments and varying ground temperature with depth. The result is that at very low flowrates the temperature could be 4-6degC  higher. However at the design flow the temperature is only 0.1-0.5 degC higher. This is because decreasing the pressure downhole increases pressure drop in the tubing and so temperature drop of the process fluid is more (which pretty much cancels out heat gained from the downhole pressure drop).

 

The temperature drop at a pressure drop downhole is less as the wellbore is in the dense phase region (300-400 barg)

 

The purpose for this exercise is as an alternative to a wellstream heater. From what I have simulated I think a heater is the way to go unless someone has some proof this downhole choke can work.  

 

Appreciate any help

 

Thanks

 

 

 

 



#2 Bobby Strain

Bobby Strain

    Gold Member

  • Members
  • 3,528 posts

Posted 10 February 2015 - 10:59 PM

Don't bet on us to give you proof. Should it fail, we're not to blame.

 

Bobby



#3 RockDock

RockDock

    Gold Member

  • Members
  • 257 posts

Posted 10 February 2015 - 11:48 PM

I've dealt with injection designs and operations for EOR, so I can only address part of your problem, but it may be very helpful. In my experience Hysys vastly over predicts methanol needed for hydrate suppression. The three primary components needed to predict hydrate formation are H2O, C1 and CO2. Of course, other components play a role, albeit, a lesser one. Hysys is notoriously bad at this prediction, especially at high pressures.

 

This link (which I just posted in another thread) shows Hysys over predicts water in the vapor phase by about 400% at your pressure. That would lead to a much greater methanol flowrate. I don't see a problem with Methanol injection at pressures greater than 100 bar.

 

http://www.jmcampbel...phase-behavior/

 

The same link shows ProMax is basically spot on in predicting the phase behavior.

 

So, while I don't know about the specific question you asked, I suggest addressing the underlying problem with the methanol injection issue at high pressures. You should be aware that almost all of the methanol injection systems are designed using ProMax. Do not depend on Hysys to give you accurate results for hydrate suppression, especially methanol injection.

 

The EOR systems I've worked on all used ProMax and the results were very good. Every single one of the EOR projects required hydrate suppression of some sort at pressures above 100 bar. 



#4 frpe

frpe

    Veteran Member

  • Members
  • 47 posts

Posted 11 February 2015 - 03:03 AM

when modeling a gas well I have found that adopting generic correlations (as those included in simulators) may produce unaccurate values for heat transfer, 
perhaps you can verify this comparing values (measured vs. calculated) and introduce some correction factor if required.
About amount of inhibitors I agree that different models may give different results,
to add my experience, for hydrates, when I compare results of Prode models vs. Hammerschmidt I see large differences in required amounts (of inhibitors).
Of course a heater would solve the problem increasing the temperature above hydrate formation point,
however make sure to calculate correctly all the properties including hydrate formation point (which may be not simple for a saturated gas at high pressure), 
as said I have a different software and no experience with your simulator but the comments from RockDock would suggest that some specific care may be required.


#5 ankur2061

ankur2061

    Gold Member

  • Forum Moderator
  • 2,484 posts

Posted 11 February 2015 - 04:59 AM

I have been reading many posts about using a specific software for a specific application and some other software not able to do it's job for that specific application. From all these discussions, it appears that software takes precedence over fundamental knowledge of engineering principles and laws. How many of those who use sophisticated simulation software know what is happening at the back-end of the software and which laws, equations and formulas is the software using? Do all of us know whether some of the default inputs the software selects itself are applicable to the specific engineering work required by us? Are we doing things mechanically by rote?

 

Some of the software discussed are pretty expensive and nobody in his or her right mind would spend a large amount of money to buy such a software if he or she has a one-time usage and no extensive future usage of such a software.

 

Per my understanding, chemical process plants were designed, built and operated successfully even before the advent of such simulation software. So to conclude that without a particular software the job cannot be done properly in my opinion is completely erroneous. Some of the finest engineering design work done in the mid-20th century was done by engineers who used slide-rules and graph sheets to generate calculations and make designs based on published engineering literature.

 

Coming to specifics, the "Hammerschmidt" equation for hydrate inhibitor injection has been advocated in the GPSA Engineering Databook. While it may not provide the most optimum injection rates, it certainly allows the engineer to have an engineering estimation for the hydrate inhibitor injection rate.

 

Some of the field engineers I know who have been involved in commissioning of process units have always told me that you as a design engineer may do whatever design calculations you like for injection rates of specialty chemicals, we as commissioning engineers always establish these injection rates based on field trials with fair amount of trial-and-error methodology. And they also say, 8 out of 10 times the injection rates calculated sitting in a design office are quite away from the actual injection rates required for trouble free operation.

 

To conclude, engineering software is not the one and complete solution for engineering design. It is the underlying principles of stoichiometry, thermodynamics, heat and mass transfer and chemical reaction kinetics that ultimately define engineering design. Let us try to remember that as engineers.

 

Regards,

Ankur


Edited by ankur2061, 11 February 2015 - 05:18 AM.


#6 RockDock

RockDock

    Gold Member

  • Members
  • 257 posts

Posted 11 February 2015 - 10:37 AM

As someone who used first generation calculators and only the tail end of slide rules, I am very happy with how technology has progressed.

 

I suppose Ankur's opinion of the simulators being too expensive is relative to the work you do. I for one will spend $10k - $100k to assist my work. A hydrate problem can shut down your entire plant, costing you millions of dollars a day. I don't want to be the one holding the bag when something goes wrong. It is my due diligence to use the best tool possible when designing or operating such a unit. The Hammerschmidt equation is helpful, sure, but significant advances in understanding hydrates have developed since that originated. I used it in my early days, and still do on occasion.

 

No one is advocating forgoing basic thermodynamic understanding. I find it very alarming when engineers are using the wrong tools for any particular problem. Knowing which tool to use for any given problem is an important part of being a process engineer.

 

A well rounded engineer will have good knowledge of of thermodynamics and the ability to efficiently apply it in a number of ways, including simulation.



#7 ChemEng01

ChemEng01

    Gold Member

  • Members
  • 84 posts

Posted 02 April 2015 - 07:09 PM

Hey guys

 

My question wasn't really to do with methanol injection prediction. I've used HYSYS and  Hammershmidt / nielsen methods including the GPSA methods for determining methanol vapourisation / losses to HC phase. Both simulator and equation methods have come up with similar results so predicting methanol injection rates is not the issue. 

 

My question is regarding the use of a downhole choke. This choke (or orifice) would be put downhole (3-4km), and do the pressure reduction downhole instead of at the surface. As the ground temperature downhole is approx 100-120 degrees Celcius the ground will put heat into the fluid which should result in similar flowing tophole temperature (FTHT) at lower flowing tophole pressures  (FTHP). So essentially making use of the high ground temperatures downhole.

 

 

For Example

 

No downhole choke: FTHP 140 barg @ FTHT 35degC

 

With downhole choke: pressure is reduced  downhole so FTHP = 100 barg @ FTHT 35degC.

 

A link to a downhole choke description is here. 

 

http://www.offshore-...rface-heat.html

 

 

Regards



#8 serra

serra

    Gold Member

  • Members
  • 310 posts

Posted 03 April 2015 - 07:05 AM

reading your first post it would seem that you were concerned about some unconsistent results from your software,

indeed as far as I know ithis doesn't seem a simple task,

consider the two steps pressure drop approach discussed in the document

An initial, roughly adjusted, gas expansion at the bottom of the well where the pressure and flow rate are high

A final controlled, gas expansion at the well head, allowing production control

 

you need

an accurate thermo to predict V,H etc.

a  specific tools to model a RO (or some equivalent device at the bottom)

an accurate evalution of heat transfer

 

I have several doubts that your software can do all these with the desired accuracy,

hence the difficult to model these processes






Similar Topics