Oct-2013

How glycols affect acid gas removal

How to control the amine unit’s full acid gas capture performance over time and so avoid operational surprises involving glycol build-up

TORSTEN KATZ, GEORG SIEDER and JUSTIN HEARN
BASF SE

Article Summary

Raw natural gas usually comes from oil wells, gas wells or condensate wells. Besides methane, it contains further valuable components, such as ethane, propane, butane and other hydrocarbons. However, unwanted components such as water, nitrogen, carbon dioxide, hydrogen sulphide and other trace sulphur components are quite common. Before sending the gas to a sales gas pipeline or before liquefaction, some conditioning is required, to purify the gas and to fulfill pipeline or LNG specifications.

Conditioning takes place in several steps and sometimes starts near the wellhead. In many applications, several wells from one or several fields are feeding raw natural gas via gathering pipelines to one central processing plant. Depending on project specifics, the length of the gathering pipeline system can consist of thousands of miles of pipes, interconnecting the processing plant to upwards of 100 wells in the area.¹ Monoethylene glycol (MEG) is sometimes injected into the gathering systems. Its high affinity towards water suppresses hydrate formation and avoids plugging of pipelines.

In the central processing plant, the final purification takes place. Here, pipeline or LNG specification of the natural gas is ensured, including the adjustment of the acid gas content and the water dew point. A common setup is shown in Figure 1.

 For sales gas applications, where the natural gas is sent to a pipeline, glycols or silica gels are common means of adjusting the water dew point. According to the Gas Processors Association, a water content of less than 7lb per million cubic feet is a recommended value for pipeline quality.⁴ For LNG applications, molecular sieves are the only option to achieve a water specification of less than 0.5 ppmv, which is necessary to avoid freezing in the cryogenic section of the plant. The dehydration unit is usually downstream of the acid gas removal unit (AGRU). For AGRUs using amines, this is always required, since the gas leaves the AGRU under more or less water-saturated conditions.

In some applications, LNG plants receive their feedstock from a common pipeline grid. This is because major LNG production facilities always require seaport access, whereas the gas fields may be located far away from the processing site. SEGAS LNG in Egypt, for example, has such a setup. This unusual pipeline/processing setup will be found more often in future: most of the US and Canadian LNG production facilities will receive their feedstock from the sales gas pipeline grid and do not use dedicated pipelines. For these cases, the receiving gas has already undergone a full conditioning process, thus the natural gas may have been processed by using glycols in the upstream conditioning process.

Another example of the use of glycols in natural gas applications is Schroeter et al’s report³ about the setup of the In Salah gas plant in Algeria. This unit consists of three pre-processing plants, in which, among others, the water dew point of the natural gas is adjusted by using a triethylene glycol unit (TEG) before compressing the gas and sending it to a central processing facility, where a second TEG dehydration downstream of the AGRU is installed (see Figure 2). 

The previous description shows that different glycols are being used at different points in the production chain, in most cases downstream of the AGRUs, but sometimes also upstream of the AGRUs.

Glycols in natural gas conditioning
The most common types of glycols in natural gas application are monoethylene glycol (MEG), diethylene glycol (DEG) and triethylene glycol (TEG). Whereas MEG is mainly used as an alternative for methanol for hydrate inhibition (direct injection into the pipeline), DEG and TEG are being used for dehydration purposes.

Due to its higher boiling point, TEG can be more easily regenerated to a higher purity; hence, it achieves better water removal and lower dew point than either DEG or MEG. A disadvantage of TEG is its higher viscosity, which can make liquid handling in plants under low temperature conditions very difficult. In these applications, DEG is the preferred glycol. The setup of a glycol dehydration plant is shown in Figure 3.

To avoid condensation of hydrocarbons in the glycol contactor, the lean glycol temperature is 
recommended to stay 10°R (5.5 K) above the gas inlet temperature.^4,5 Common feed gas temperatures between 5°C and 50°C (41-122°F) result in lean glycol temperatures between 10.5°C and 55.5°C (51–132°F). Taking a usual TEG flow rate of 6–8 US gal/lb H2O into account,⁵ the resulting heat of water condensation for common natural gas pressures (p = 40–75 bara, 580–1088 psia) will only marginally contribute to an increase in the feed gas temperature (usually <3.6°R or 2 K). Lower feed gas pressures may lead to somewhat higher temperatures in the treated gas. However, the assumption that the glycol contactor dry gas temperature equals the feed gas temperature allows the approximation of the minimum glycol content caused by vapour pressure losses in the treated gas.

In contrast to amine absorbers, a glycol absorber can never have a water backwash section to reduce vapour pressure-induced solvent losses, since the gas would immediately saturate with water again. As a consequence, the treated gas exiting the glycol absorber will always be glycol saturated for the pressure and temperature conditions in the glycol absorber top (see Figure 4).

Even though glycol vapour pressures are very low, a low concentration of glycols will always end up in the glycol absorber treated gas phase. Losses by entrainment will further increase the content. Table 2 shows two gas compositions, which will be used for a case study in this article.

Case 1 represents a CO2-rich gas, which is also rich in C2 and C3+ components, whereas Case 2 represents a leaner gas, coming from a pipeline.

Figure 5 shows the glycol saturation concentration over the temperature range for gases with compositions according to Case 1 and Case 2 for pressures of 50 and 70 bara (725 and 1015 psi). The values have been determined by using the commercially available software Multiflash,7 using the cubic equation of state PR (advanced). Even though the gases are quite different with respect to acid gas content and heavy hydrocarbon content, neither of these two parameters has a substantial impact on glycol solubility, at least not for typical LNG feed gas conditions.