three different quotes: a typical TEG plant, a TEG plant with a Stahl col - umn, and a traditional molecular sieve dehydration process. All three processes were based on the same feed used in the ProTreat simulation. Prices are quoted with a West Texas freight-on-board basis of current prices. The final cost estimation of each process is shown in Table 2 . Analysis shows that the typi - cal TEG and Stahl processes have similar cost estimates. The main equipment difference in the quoted processes was that the TEG reboiler and BTEX condenser in the Stahl process were actually sized smaller. However, the Stahl process had a slightly larger column diame- ter. Since the Stahl process had an extra (Stahl) column, the final cost was slightly higher than the typical TEG process. However, the molec - ular sieve process was much more expensive than either TEG based process. Overcoming pitfalls In this study, the ProTreat simulator clearly demonstrated the possibility of achieving very deep dehydra - tion using a Stahl column. Through modifying the stripping gas flow rate and Stahl packing depth, it was found that the dehydration level was sensitive to design. Despite these encouraging results, there exist several pitfalls. When design - ing these systems, the dry gas and hydrate equilibrium curve must be evaluated because very low water concentrations can be conducive to hydrate formation. Additionally, TEG tends to absorb chemicals besides water, such as BTEX, amines, and oil, which can each lead to fouling. Keeping the Stahl column packing free from fouling is important to prevent long-term maldistribution, which is known to be a problem in at least one facility. There are also the nor - mal process risks of entrainment, leading to the potential presence of TEG in the dry gas and other issues such as excessive reboiler tempera - tures causing thermal decomposition and vaporisation of TEG. Aerosol and even vapour tail TEG can also form an epoxy-like material, leading downstream to process disruptions.
10
Bestani-Shing Parrish
1
0.1
0.01
0.001
0.05
0.1
0.15
0.2
0.25
0.3
Stripping gas rate (MMSCFD)
Figure 2 Effect of stripping gas flow rate and VLE model
Typical TEG
Stahl TEG
Molecular sieve
Total price, million USD
2.32
2.75
10.89
Table 2
dicts a higher water content by a factor of 10 than the Bestani-Shing model at the original simulation stripping gas flow rate of 0.16 MMSCFD. However, it has a lower predicted water content at higher stripping gas flow rates. In the second test (see Figure 3 ), the Stahl column packed depth was varied at a constant stripping gas flow rate of 0.25 MMSCFD, as both VLE models showed a similar dry gas water content at this value. This test shows that at greater column heights, a very low water content, below 0.01 ppmv, can be achieved. The Parrish-Margules model pre - dicts a lower water content at higher packing depths but a higher water content at lower packing depths.
These results show how modify - ing the stripping gas rate and the Stahl column’s packed height affect the dehydration level significantly. Furthermore, there is a difference between VLE models that can make dehydration levels differ by an order of magnitude. When design - ing a Stahl column dehydration unit, the difference between models might be viewed as a rough meas - ure of the margin of error. Economic evaluation To evaluate the economics of the Stahl column process compared to a typical molecular sieve dehydra - tion unit, an independent cost esti - mation quote was obtained from Reset Energy LP. Reset provided
10
Bestani-Shing Parrish
1
0.1
0.01
0.001
2
5
8
11
14
Stahl column packing depth (m)
Figure 3 Effect of Stahl column height and VLE model
24 Gas 2022
www.digitalrefining.com
Powered by FlippingBook